5 things you need to migrate web analytics on-premises to SAAS

netinsight sunsetRemember the IBM announcement back in April to sunset NetInsight? The truth is on-premises web analytics is a dying art. The only other well-known vendor that provides on-premises solutions is WebTrends. It is destined to suffer a slow death – there have not been any new releases for the last few years and customers are encouraged to move to the SAAS version. So, the next best thing is to prepare and plan for inevitable – migration to SAAS.

  1. Assess Web Analytics vendors. Sunsetting the on-premises solution presents a good opportunity to reassess the web analytics landscape instead of blindly sticking with the same vendor and moving your data from on-premises to SAAS.
  2. Documentation. Documentation. Documentation. I said it three times and I meant it. No one likes to create it. I get it – it’s a boring, monotonous task to write every variable, every processing rule, every customization. However, switching your web analytics tool without documentation is like going into battle and forgetting your ammunition. You need to ensure you have documented the following:
    • A List of Key Custom Built Reports and Layouts. This is your foundation for stakeholder expectation management and therefore the golden key to your sanity during migration. This task can become a project on its own since going through such an exercise will confirm what reports and metrics are critical and important for the business and which ones you can delete because no one is using them anyway.
    • Current Tool Configuration rules. This is a big one and can cost you dearly. I had many urgent calls from clients desperately trying to understand why their data tanked 20 – 30% when they switched to a new tool or data collection method. In 99.9% of all cases the answer was due to configuration such as page view definition, filtering, visitor tracking methods, etc. The configuration topic certainly warrants a separate post, so check back in a while.
    • Site/Metric Matrix. List all custom collected metrics per site, including metric definition, collection method and syntax. This exercise should be completed hand-in-hand with report documentation to ensure every single custom metric is documented. This is your bible. Keep it on your nightstand and refer to it often. If you need a sample template, come back in a few weeks – I will be posting further on this subject.
  3. A Project Manager and a Project Plan. Web Analytics migration is like surgery – you need to make sure that your patient (aka reporting) will not die during the surgery (migration). You may get away without a plan if you a dealing with a simple site and basic data collection (wart removal), but if you a dealing with multiple sites, channels, applications and vendors (open heart surgery), a solid plan is required for successful execution.
  4. Assessment against Current Reporting Capabilities. Migration from on-premises to a SAAS solution (or one vendor to another) almost always results in loss and/or gain of features. E.g., SAAS solution is likely to have greater social and mobile tracking capabilities but you may need to make some data collection tradeoffs in order to adhere to your company’s data collection policies, especially if you are in the healthcare or financial industry. Create a loss/gain matrix and use it to manage change. No one likes to give up things that they already have. Over communicate any changes to stakeholders in advance, quantify impact of such changes on business and help to define mitigation plan, if necessary.
  5. Assess Current Roles, Responsibilities and Processes. Switching from on-premises to SAAS will impact current roles and processes, e.g., there will be no need to maintain servers.   Process-wise, you will not be able to re-analyze collected data, which will have an impact on new report deployment as well as the ability to apply new reporting requests to historical data. Don’t wait for a bear to get you – review current roles and processes, outline necessary adjustments and manage change before the migration occurs.

Feel free to comment or add to my list!

The Consumerization of Health Insurance: Adapting to Private Exchanges

findcustomersThe new Affordable Care Act (ACA or ObamaCare) is introducing new opportunities and challenges for health insurance companies. The complex set of regulation, exchanges and integrations needed is still a political and technical mess but one thing is clear: health insurance companies will have to embrace their consumers.

It’s no secret that currently most health insurance companies’ customers who make buying decisions are not the actual consumers but employers or benefit brokers. This is about to change.

The process started even before the latest reforms and is modeled to a large part after the successful pension / retirement benefits model where companies moved from a company provided pension into a marketplace. The employer is putting in a defined contribution, and the employee is choosing investment vehicles from various providers.

The model works very similarly for health benefits with private exchanges giving employees more choices. Walgreen has recently moved its 160,000 employees to a private exchange and estimates are that by 2017, 18% of the American public will buy their insurance at private exchanges.

So what do health insurers need to do to better compete in an open marketplace? Mostly, steal the best practices established by other competitive markets such as the aforementioned retirement benefits and P&C insurance providers:

  1. Enable consumers to make an informed buying decision. While prices and coverage may be negotiated with employers, additional tools and content written for consumers is essential. For example, a “find a doctor” tool that lets you see if your physician is part of the plan, and detailed coverage comparison between plans.
  2. Give consumers full access to their information and personalize the experience. Web portals, mobile applications, email and text messages all tailored to consumer preferences and health interests. The self-service aspect will both give consumers control and save call center costs.
  3. Know your customers. Since until now most group members were not customers, the interaction was very transactional and focused on claims. A huge part of the consumerization of healthcare and of health insurance is starting to use ecommerce style tools – CRM systems that help track and manage all interactions; improving data collection, tracking and analytics to help segment and personalize user experience and wellness communications and offerings.
  4. Establish a clear measurement and analytics framework. New measures and metrics need to be put in place to judge the effects of this transition on the business and the determine best ways to react. The new measurement framework has to look at metrics such as:
    • Customer acquisition cost
    • Customer retention rates 
    • Customer profitability by source and segment
    • Customer lifetime value
    • Impact of wellness activities and user engagement with them on costs
    • Impact of self-service portal and mobile applications on call center volume and costs.
  5. Adapt and optimize marketing. The direct approach requires multichannel direct marketing.  Analytics can help guide the best mix of marketing options to achieve the different acquisition and retention costs.

A lot of people question whether going after the direct channel is even worth it. Some have had bad experiences in the past with individual members that tended to consume more healthcare since they were not always in a good enough health to hold a full time job.

The transition we are seeing to private exchanges and defined contributions seems much more substantial and can dominate the market in 5-10 years. A good toolset, marketing approach and measurement framework will be invaluable to compete for the right segment.

From Web Analytics to Customer Intelligence

CIWe recently were invited to present internally at a prominent health care payer network about the rapidly changin role and importance of web analytics. Gone are the good old days when it was enough to just run a log analyzer or put a simple tag to collect all the information needed about the interactions a customer has with you. Analysis used to be limited in scope and focus on a handful of parameters that could be optimized, such as bounce rates and conversion rates, by tweaking the checkout flows and usability improvements.

Not that conversion rate optimization is less important today but as customer interactions focus less and less on just the company website, the new critical need is to try and get a coherent picture of general customer behavior across all touch points. Instead of trying to infer customer thoughts and concerns through their clickstreams, many are now openly expressing needs and problems through social media.

This goes beyond “cross channel marketing” into the new area Forrester and others are now calling Customer Intelligence (CI). Similar to the way business data evolved from simple reporting into Business Intelligence (BI), as customer data gets more complex and varied, putting everything together and drawing conclusions and trends from it will need to employ similar methods and tools.

This is primarily a mindset change from the somewhat passive “analytics” to the broader and much more active role of managing and providing customer intelligence.

The expectations from Web Analytics professionals and systems are changing as well from the cyclical analysis and response to the providing of on demand, immediate intelligence for both individual and aggregate customer needs and problems. In some companies this evolved into a real “command center” that has 24/7 monitoring and interaction tools to listen, interact and respond to customer needs.

There are a few challenges that mark this transition:

  • Quantity: The quantity of interaction points is exploding due to social media, online videos and mobile devices.
  • Traceability: It is very hard to identify users across various media. Mapping a web user to a Facebook account or twitter feed is not always possible.
  • Immediacy: There is an overwhelming need and expectation for immediate response.

Here is a conceptual diagram of this new reality illustrating all the new interaction points being consolidated into the central Customer Intelligence and the introduction of the analytical services that can be used to optimize the user experience.

These analytical services can work on both an individual and aggregate level:

  • Individual: If we can aggregate customer data and interactions from different channels, this will dramatically improve segmentation, insight for sales and customer service professionals interacting with the customer, and services that can target offers or content in real time based on user past interest and behavior.
  • Collective intelligence: By looking at customer activity across all channels we can:
    • Optimize targeting through the different channels and our investment in them
    • Improve recommendations
    • Identify trends
    • Identify problems / issues / sentiment changes and address them quickly.

To start implementing Customer Intelligence, the process is now becoming quite similar to implementing a BI solution

  • Expand use of social listening and data capturing tools and store their data
  • Adjust data models to accommodate multiple user identifiers, channels, devices etc.
  • Redefine KPI’s
  • Define and implement analytical services
  • Adjust reporting and analytics
    • Real time
    • Dashboard level

The Web Analytics vendors are starting to step up and offer tools and support for Customer Intelligence. In upcoming posts we’ll look into WebTrends, Omniture, Google and IBM to see how their offerings stack up and the type of solutions they support.

Adobe, IBM, WebTrends, and comScore named leaders in Web Analytics

Independent research firm Forrester recently released their annual “Forrester Wave: Web Analytics, Q4 2011” report naming Adobe, IBM, comScore, and WebTrends as the current leaders of the web analytics industry. AT Internet and Google Analytics were also included as “strong performers” while Yahoo Analytics took 7th place as the lone wolf in the “contender” category.

Not surprisingly Adobe Site Catalyst and IBM Coremetrics stood out with the top two scores overall but WebTrends Analytics 10 and comScore Digital Analytix showed major stengths as well. Unica NetInsight, another offering from IBM did not make the list because of its inevitable fate to be merged with Coremetrics. In 2010, IBM acquired both Unica and Coremetrics. The Forrester report states, “IBM is incorporating the complementary and notable features of Unica NetInsight into a merged web analytics solution based on the Coremetrics platform.”

The full report can be downloaded from Adobe or WebTrends and will likely show up on other vendor sites soon.

Multi-Touch Attribution Campaign Tracking with WebTrends

This article is a follow-up to the webinar

All web analytics platforms have some way of tracking marketing campaign performance usually out-of-the-box or with a little bit of set up. Generally they all do a pretty good job of this and provide key reports to make important business decisions about which campaigns to invest more money in, which to reduce spending on, and which to get rid of altogether. But often these decisions are made without insight into the whole picture. Why? The answer is simply because most campaign reports are set up in the industry standard way of attributing all conversions to the last or most recent campaign clicked. This is and has long been the industry standard, but it is time for a change as this method ignores the fact that people often go through multiple campaigns before converting.

So what other attribution options are there? And why wouldn’t I want to attribute conversion credit to the most recent campaign? – There are typically 3 options for campaign attribution:

  1. Last Touch (Most recent campaign)
  2. First Touch (Original campaign)
  3. Multi-touch (All campaign touches)

Technically there are two options for multi-touch attribution. One option is to give full credit to all campaign touches and the other option is to give partial credit to each touch. For example, if 3 different campaign touches resulted in a sale of $30 you could credit each touch with $10. But for the purposes of this article we will focus on the full credit option. As for the question “why wouldn’t I want to attribute conversion credit to the most recent campaign?” – this is not really the right question to ask. The better question to ask is, “Do I have the best possible insight into the performance of my marketing campaigns?” The answer to that question is almost always “no” if you are only analyzing a single attribution method. So rather than replacing industry standard last touch reports, adding first touch and multi-touch to your arsenal of reports is the best course of action.

Fortunately for WebTrends users, there has been a great method for gaining insight into all campaign touches for quite some time although a little work up front is necessary to gain the full power of this. If you are already doing basic campaign tracking within WebTrends then the visitor history table is already turned on and with minimal effort you can set up two new custom reports which report on the first touch campaign and all campaign touches respectively. To do this you need to make use of two features of the visitor history table and create two new custom dimensions, one based on WT.vr.fc (the fc stands for “first campaign”) and another based on WT.vr.ac (the ac stands for “all campaigns”). Once you have the dimensions set up you create custom reports using those dimensions and whichever metrics you want applied. To make things easier, copy the existing campaign ID report and just change the dimension to base the report on.

The “first touch” report ends up looking nearly identical to the existing campaign ID report but the rows of data will be different since the revenue and other conversion credit is applied to the first campaign that referred the conversion as opposed to the last.

Standard Campaign ID Report Sample
First Touch Campaign ID Sample

The “all touches” report is where you’ll notice more differences. You will see some or many (depending on the date range you have selected) rows of data that have multiple campaign IDs separated by semi colons. To view only the data that contains multiple campaign touches just filter the report by a semi colon.

Multi-Touch Campaign ID Report Sample

So what do you do with this information? What does it all mean?
Spending some time with this new data will likely reveal some patterns you never had insight into before. For example, you may notice certain campaigns appear to perform poorly according to your traditional last touch reports but the same campaign’s performance as a first touch is much better, or vice versa. Since the first touch report is so similar to the out of the box campaign ID report it is fairly straightforward. The only difference is that the first touch gets the credit. The all touch reports are more complicated though. What I find most useful about this report is the ability to determine a campaign’s total reach and compare it to its absolute reach.  Take for example campaign ID 32. In the above screenshots you will notice that this campaign ID has $63,441 attributed to it as a last touch campaign, $35,839 attributed to it as a first touch campaign, and $82,036 attributed to it when you search for it in the all touches report (See fig. 4 below). What this data is telling us in this particular case is that:

  • $63,441 in revenue was most recently referred by campaign 32
  • Only $35,839 in revenue was initially referred by campaign 32
  • But overall campaign 32 at least partially referred $82,036 in revenue

As you can see, there can be very significant differences in campaign performance depending on how you look at the data. Taking the easy way out and looking only at a single attribution method can lead to less than fully-informed decisions being made about your campaigns. What if you were relying solely on first-touch reports in this example? That could lead you to reduce your budget on campaign 32 when in reality it was performing much better than your first-touch report told you.

Multi-Touch Report Filtered by Campaign ID 32

Ok, so all that is well and good but manually analyzing campaign IDs one at a time is a lot of work! Yes it certainly is using the methods I just provided as examples. But there is a much better way to approach this. Taking things a step further we can export each of these reports and combine them together in Excel using the campaign IDs as our key values. What we want to end up with is something like the following which will allow us to analyze first, last, and multi-touch all within a single interface.

Multi-Touch Reporting in Excel Sample

In part two of this article I’ll show you how to set this all up in WebTrends. But for now, follow the steps discussed in this article to get these super handy reports in place so you’ll be ready for the next part.

Your Company’s Social Debut

Planning Your Company’s Debut or Strategy in the Social Media Sphere

Corporations have long been regarded by the law as having “legal personality”-  which means they have rights, privileges, responsibilities, and protections just like humans (with some differences, like marriage).   It should come as no surprise then, that they’re acting like humans more and more – now they’re relaxing with friends, and socializing! As communication gets easier through digital technology, humans are now able to interact with corporate personalities.  And these personalities are just beginning to awaken to the new freedoms they can find in the digital landscape.

If you’re like me, and I bet you are, you are both human, and, also a part of bringing business personalities to the social scene. In this capacity, I recently attended SocialTech2010 in Jan Jose, CA, right from my desk in NYC.

As the Twitter stream flowed by rapidly with commentary and quotes from the speakers, I watched and listened to advice, case studies and stories from the experts on Social Media for Business. I came away with the recognition that Social Media for business is just like a big networking cocktail party!

Companies aren’t accustomed to acting as social creatures and the adjustment will take some time. We all had to learn social skills growing up; companies can do the same. There are a few things that etiquette would require of a cocktail party attendee and that’s the same strategy the speakers at SocialTech2010 are recommending:  Know who you are, be interactive and respectful, don’t gossip, be a good listener, and don’t be afraid to share yourself.

As businesses gain proficiency in this kind of interacting, they follow an arc towards maturity. Kathleen Malone of Intel outlined the following 5 stages of a Social Media Approach:

1)      Listen: In this stage a company finds out: What are people saying about my Brand and/or my field? Where are they having this discussion? Who are the major players and influencers?  Services like Radian6, which Malone says Intel deployed 18 months ago, make this possible.

2)      Analyze: This is the time to read the room/space, figure out what your angle will be when you eventually do pipe up. Which conversation will you enter? What are your expectations? Why are you going to participate?

3)      Create: This is the stage where the business comes up with something appropriate to say. To participate effectively in the conversation, Malone says your content should be: useful, interesting, human, “snackable” (meaning in bite size pieces, easily consumed), inspiring and should cater to egos and build community.  

4)      Engage: In this stage you go public and enter the conversation, getting your content out there in new ways and/or by participating in the conversations that already exist.

5)      Measure: Your social media approach is not complete without an understanding of how you’re doing. The internet is an amazing forum for measuring how people behave with your content, and you should use a variety of tools to understand the response to your forays. Measuring properly will provide insight on how to proceed, both in the ongoing conversation, and with the business itself.

Both Malone and Brian Ellefritz of SAP outlined the natural evolution of Social Media programs at large companies  – first there are what Ellefritz calls “Grass Roots” efforts, where excited individuals branch out in ways that are unpredictable and non-uniform. He says companies should encourage these exploratory missions. Leadership will begin to emerge internally, and informal education will get the ball rolling. Following the “Grass Roots” period, Ellefritz sees “Silos Form.” This may not feel 100% smooth, but is an important step, as “coop-eteition” (a kind of cooperating/kind of competing relationship, sort of like sibling rivalry that spurs each one on) sees different silos jockeying for position. During this step, Ellefritz encourages companies to “invest in leaders, not laggards”, and to get the players from various silos together to learn from each other.  Also, he says, “don’t wait too long for governance.”

The next evolutionary phase in a corporate Social Media Program is “Operationalizing” – where leadership becomes clear, channels become well formed and in alignment with the divisions in your business.  Tools begin to consolidate and more emphasis on measurement and results appears. By this point your business may have headcount devoted to social media, and content should become less problematic, less of a focus, because it’s running more smoothly.  During this stage it’s important to align and integrate silos, and focus on strategy, ownership, metrics and priorities.

After this shift, the next phase is what Ellefriz calls “Lifestyle.” This is when the Social Media program has engaged and competent employees and success is understood and positive outcomes are frequent. This is a level of Social Media implementation that is fairly rare in today’s scene, though Ellefritz points towards Zappos as an example of a company that may be at this level.

.. .. ..

The wonderful thing about participating in social media is that it lets your personality out! For a business that hasn’t previously seen itself as the kind of entity that has a social life, this might seem daunting at first.  That’s why Ellefriz’s evolutionary arc makes so much sense to me. The way I see it, people and businesses want more than ever to get clear on who they are, and who they want to be, in order to present themselves well, and to participate in Social Media conversations. The best advice is to be authentic. Just like at cocktail parties, the people you’re conversing with generally know if you’re “full of it”, or if you’re being sincere.  Your conversational counterparts like to be complemented, offered nuggets of useful information, and generally considered and included.

For businesses, (and the teams of people that perpetuate them) this will mean really focusing on what the goals are, what opportunities exist to communicate clearly and uniformly around these interests, finding “friends” out there to talk with, and owning up to the inevitable minor mistakes that are so easy to make along the way. Since SM is such a public sphere, the resulting increased level of transparency is going to make businesses change and open up in new ways.

Coachdeb:”RT @MarketingProfs: “When someone says they need a Facebook strategy, a Twitter strategy, I say… Wait! Take it back… What’s your story?” @scobleizer #mptech”

So, armed with the Social Media/networking party analogy and with the stages of approach and evolution path laid out before you – what are you waiting for?  Participate!

Here are 10 tips to consider as you get started:

1)      Go where the fish are – target engagement carefully where the conversation already is.

2)      Social Media is Local. The goal is to be uniform while being decentralized – Intel communicates internally with their 1000 “Registered Social Media Practitioners” with guidelines and trainings (some mandatory). Intel also has their own internal newsletter that aggregates Social Media content – Malone says this makes management comfortable as well as keeps everyone updated.

3)      Have a Content Calendar for the year to coordinate Social Media messaging across channels and people, and to keep it focused on your message. Kathy Malone said at Intel, 2/3 of the content that gets put out falls under the guidelines of their content strategy calendar.

4)      Consider in advance how to manage Social Media Risk. One of the most interesting things Jaime Grenny of SalesForce said at SocialTech2010 is that all their employee training videos on Social Media strategy (and how to use online video for B2B marketing) are up for the public to see on YouTube (here).  This level of transparency lets everyone know what to expect upfront.  Malone outlined a “prevention/detection/response” approach in which 3 teams worked from different angles to mitigate risk on the social media front. And experience teaches: “if you screwed up, fess up”, and be transparent.

5)      If your company is doing moderation of dialogue, consider having a light hand to keep the conversation honest – as Intel puts it, they let the good and the bad in, but moderate the ugly – mostly meaning profanity and non-constructive comments, and they’ve found their audience appreciates it.

6)      Build a business case for your business so you know why you’re entering into Social Media – not only will it legitimize your efforts internally, but it’ll provide clarity for your message. Will it extend customer service? Will it increase SEO? Can you use it to create brand advocates and champions? Can you collect ideas on where to take your product?

7)      To measure, use Context. As with all web metrics, in order to understand what’s happening you need to understand the context of your data, and compare it to a baseline to view trends. Knowing your goals will assist you in setting up context.

8)      People are the PlatformLaura Ramos of Xerox encourages us to get our people out there and seen. Show video of your thought leadership. Get your salespeople to share their stories and knowledge with the rest of your company and make them heroes. Build relationships, and let your existing customers create new business for you. Social Media Marketing is not about reaching many to influence a few but engaging a few to influence many!

9)      Social is relevant. Here are some StatsRené Bonvani of Palo Alto Networks says that FaceBook has a 96% penetration in enterprise, meaning that only 4/100 people aren’t using it at work! He also said that only 1% is posting on Facebook but that people are 69 times more likely to use FaceBook chat than to post.  Another impressive Bonvani stat: 69% of business buyers use social media to make purchasing decisions.  No matter the numbers, it’s clear that with the cost of communication dropping close to $0, as social beings, we’re using the web to communicate more often with more people, and in smaller chunks regularly.

10)   Social media has to be part of WHAT you do, not something else you do. Jeremiah Owyang in his keynote said that the only difference between the Social Site and your business is the URL. He says that in the radical future, websites will be dynamically assembled on the fly based on social profiles. URLs and domains won’t matter – the web will be sorted around people and contextual situations.  Because of this, ads will become useful content.  This is already evident.

So – Get out there and participate!

Edgewater Technology provides strategy, consulting, web metrics, and implementation expertise to help you focus on the best ways your company can engage in these dynamic communities and track your success!

10 Actionable Web Metrics You Can Use – Part 2

Show your analytics results with gauges

In Part 1 of this post, I discussed 5 percentage-based metrics that can provide actionable insight. In Part 2, I will go over 5 index-based metrics that can also provide insight to problems that may need to be addressed in order to maximize the value of your website.

1. Campaign Quality Index (CQI)

This index measures how well targeted your campaigns are at driving qualified traffic to your site. Suppose 40% of your traffic comes from a particular campaign, but the traffic only provides 20% of your overall conversions. The CQI for this campaign would be the percent of conversions from the campaign (20%), divided by the percent of visits from the campaign (40%). A value of one means that a visitor from this campaign is as likely to convert (purchase, sign up, request information, etc…) as from any other campaign. A value less than 1.0 means they are less likely to convert, while a value greater than one means they are more likely to convert. If the value is less than 1.0, then you need look at the reasons. You can break this down to individual search engines, or even keyword groups for each search engine, and for each individual banner campaign or other paid campaign you use, including referral partners. Perhaps the targeting is not sufficiently narrow, or the message is not being carried through the site (high bounce rate). You will want to work with your SEM team and landing page design team to make the needed changes. When you make improvements, you can track their effectiveness by watching the index change. Ideally, your analytics dashboard should be created so that you can see the changes over periods of time.

2. New Customer Index (NCI)

This index is focused on transactions (not revenue) from new customers. It is defined as the percent of transactions from new visitors divided by the site percentage of new visitors. For example, if 40% of your transactions are from new visitors, and 60% of your traffic is from new visitors, your New Customer Index is 0.67. A value of 1.0 means that a purchase is equally likely to come from a new or returning customer. A value less than one (as in this example), means that a new visitor is less likely to become a customer. A value greater than one means that a new visitor is more likely to become a customer than a returning visitor. Your goal is to strive for a value of one or better. If the value is less than one, you will need to look at factors that contribute to a low value. To do this properly, you would want to create a New Customer Index for each type of campaign you run, and compare that to those who come to your site from direct entry. A low performing index for paid search or banner campaigns can mean that you are not targeting the correct market, or that your search terms are not correlated to those looking to purchase your product or service. If the campaign is a banner campaign, either the message is not on target, or the media partner you are using is not attracting the correct demographic.

3. Return Visitor Index (RVI)

This index is simply defined as the percent of return visitors divided by the percent of new visitors. A value of 1.0 means that your site has an equal distribution of new vs. return visitors. A value greater than 1.0 means that your site is more likely to attract return visitors, while a value less than 1.0 means your site is more likely to attract new visitors. Depending on your type of site and your effort on attracting new visitors or keeping existing visitors, you can see how effective your efforts are and can then focus on how to improve this index. If your goal is to encourage repeat visits, then you need to be concerned with how fresh or relevant your content is, or how effective any email campaigns are in getting registered visitors to come back to your site. Any anomalies need to be investigated. As an example, I once saw a huge jump in new traffic in a client’s site that was the result of an email campaign, according to the analytics report. However, the email campaigns were only to registered visitors, so in order to have received the email, you would have first had to have visited the site. Thus, the email campaign visits should show up as return visitors. What happened is that the email contained an offer for a free exercise DVD, and the link URL was hijacked and placed on a few deal sites. When visitors clicked on the link, they were attributed to the email campaign, as the link contained the email campaign code! By looking at the RVI, I was able to see that there was an issue that needed to be addressed.

4. Branded Search Index (BSI)

Organic search can consist of generic terms that relate to content on your site plus searches that include your company name or your brand name.  Each can be of interest to your search manager. If more visitors come to your site from generic keywords or terms, it means that your site is well optimized for content. If more of your search visits come from branded terms, it means that more people are finding your site by your brand name instead of from non-branded terms.  You can track this by creating a BSI metric. This is defined as the percent of visits to your site from branded terms divided by visits from non-branded terms. Values greater than 1.0 mean that you are getting more of your traffic from branded terms, while a value less than 1.0 indicate that generic terms are winning the organic search battle. Depending on your search strategy and goals, you can use this information to help adjust your optimization or brand promotional efforts.

5. Site Search Impact (SSI)

Site search is very important for many types of sites. Visitors who come to your site may use site search to help them quickly find what they are looking for. If they find what they want, they may be more likely to continue to reach a goal, such as a purchase or lead submission. If they don’t find what they are looking for, they may just leave the site. The SSI index can tell you the impact your site search has on your revenue. To calculate it, take the per visit revenue from those who use site search, and divide it by the per visit revenue of those who do not use site search. “Per visit” revenue is defined as the total revenue or lead value for the month, divided by the number of visits. If your SSI index is greater than 1.0, this means that your site search is making you money, compared to those who do not use search. If the index is less than 1.0, it means that your site search is costing you money, meaning those who use site search are less likely to either make a purchase or become a lead. This can be the result of not getting desired results from the search, or result pages that don’t satisfy your visitors’ needs. To solve this problem, you would then need to dive deeper into your site search report to identify and correct the issues.

Summary

Hopefully this two-part post on 10 actionable web metrics you can use has given you some insight into how to make your web analytics program more actionable. While some of these metrics are fairly easy to construct, others may require filtering, segmentation, calculated metrics and integration with offline data. Depending on your analytics tool, you may want to use a presentation package like Xcelcius to create and display your gauges and create a dashboard that can be shared with your site’s key stakeholders.

10 Actionable Web Metrics You Can Use – Part 1

Make your web analytics actionable

The end goal of a web analytics report should be to provide some guidance on how to take an action to improve how your website is meeting its goals. However, many analysts simply generate canned reports using their analytics tool and send it to their management for review. In this two-part post, I will share with you 10 different web metrics that can “at a glance” tell your management how well a particular campaign or goal is performing, plus provide some relevant actions that can be taken to improve the underlying performance of the metric.

In Part 1, I will look at five metrics that are expressed in percentages. In Part 2, I will look at five metrics that are expressed as an index. Ideally, these metrics would be designed to be seen as gauges on a dashboard, and some can have the ranges color-coded (green/yellow/red) to quickly show the impact of that metric. Here are the first five actionable metrics.

1. Campaign Margin.

If you are running any paid campaigns for an ecommerce site or lead generating site, you need to know your margin. In simple terms, your campaign margin is defined as your revenue from a campaign less its cost, divided by the revenue. Your goal is to stay as close to 100% as possible. You can create a report that shows the campaign margin for any campaign that involves external spend (banners, paid search, sponsorships, etc…), or an internal spend on employees’ time (social media marketing, forum and article posts, etc…). The smaller your margin, the less money you are making. With this metric, “0%” is breakeven. If you have a negative margin, you are losing money on that campaign. If you have a positive margin, you are making money. This type of margin can be shown as a gauge and placed on your analytics dashboard. If your margin is negative or near zero, you need to take action to look at why the campaign is costing so much or how you can increase the campaign’s effectiveness.

2. Percent Revenue from New Visitors.

This metric tells you how likely visitors are to order from you on their first visit, compared to ordering on successive visits.  In order to create this metric, you need to be able to segment your traffic by new vs. repeat visitors. To calculate the metric, take the revenue generated from new visitors and divide it by the total revenue.  If the percentage is more than 50%, you get more of your sales from first time visitors, If it is less than 50%, you get more orders from repeat visitors. If you see this percentage is low and you have limited repeat buyers, then perhaps you would want to do a better job to get a visitor to purchase on their initial visit. If you have a low percentage of revenue from new visitors, and you have a more expansive product line, then this metric is telling you that you get more of your sales from repeat visitors or customers, and you may want to focus on keeping your content fresh and maintaining campaigns such as email or social networking to keep your visitors coming back.

3. Engaged Visitor Percentage (EVP)

This metric is defined as the number of visits that contain an action or event that indicates engagement divided by the total number of visits. To use this metric, you must first determine what defines an engagement. This can be any of the following – visit a specific number of pages, visit particular pages of interest, subscribe or register to something on your site, post a comment, rate something, click on an ad, use a tool, navigate a map, download something, play a video, forward to a friend, or do anything else you wish to show engagement. By monitoring this metric over time, you can determine if your site is doing a better or worse job of engaging your visitors, if this is one of the goals of your site.

4. Utilization Factor (UF)

Some types of organizations have developed their website to encourage its users to conduct business through it instead of calling or submitting paperwork. For example, an insurance company may want claims to be processed via the web. A financial agency may want its brokers to process transactions via the web instead of sending in forms. If one of your goals is to encourage the use of your site to accomplish tasks, one way to measure this is to track the percentage of activities that are conducted on the web divided by the total number of activities conducted online and offline. This metric is a bit more complicated, as to do it entirely online you need to import the offline data into your web analytic program. You can also export the online data and create an Excel-based report that combines the online and offline data. Your UF can also be used to measure the percent of registered users who use the site to transact business. By monitoring the Utilization Factor over time, you can determine how well your efforts are to shift your transactions to the web. Specific actions can include training of your users on how to use your site to process transactions, or ongoing communications that remind your users to use the site.

5. Self Service Factor (SSF)

If your site is to be used to provide customer service, one of your goals could be to reduce the percent of customer service issues that are handled through the phone. Thus, the SSF would be calculated as the number of service issues that were resolved on the web divided by the total number of service issues (web + phone + chat + email). In order to do this, you would either need to import your offline data into your web analytics program, or export your online data into a spreadsheet to combine it with your offline data. If your company has a target goal for resolving service issues via the site, you can create a gauge that shows how well the actual percentage is compared to the goal, or color-code the result as red or green to show if the SSF is above or below the target. Part of your site’s optimization efforts would include analyzing the issues that are most often called in and updating the content on the website, or making the top 10 most frequent issues a sidebar on the customer service site.

In Part 2 of this article, I will show you how to use these five additional actionable metrics:

  • New Customer Index
  • Campaign Quality Index
  • Return Visitor Index
  • Branded Search Index
  • Site Search Impact

Is Your Web Analytics Program on Solid Footing? (Part 2)

pillars of support for your web analytics platform

In Part 1 of this topic, I covered four of the top ten fundamentals in building a strong web analytics platform. In this post, I will discuss the remaining six pillars.

5. Develop Actionable Campaign Tracking

In a previous post, I talked about tracking all of your campaign activity. A campaign is any method, whether paid or organic, that gets visitors to your site. Some of these activities include pay-per-click, banner ads, email, newsletters, blogs, articles, social media, classifieds, forums, referral partners and affiliates. In the other post, I provided recommendations on how to set up Google Analytics and Omniture to provide you with a methodology to create and track the performance of all of your campaigns. When done properly, you can determine how well these campaigns do in bring not only visitors to your site, but qualified visitors who become customers or leads for your company. Once you know the value of your campaign efforts, you can provide recommendations on which campaigns work and which ones do not, letting your organization optimize its marketing budget.

6. Evaluate Your Data Quality

The expression “garbage in, garbage out” applies to your analytics program. If the quality of the data you are processing is suspect, the quality of the reports will not be any better. Some of the items you need to pay attention to include:

  • Filtering of internal and development partner traffic
  • Exclusion of images, spiders, bots and external site monitoring services from being counted as visits and page views
  • Merging together same pages with different URLs (case differences, “www.” vs. no “www”,”/ index.htm” vs. “/” at the end of a home page or path)
  • Removing query parameters from same page names
  • Testing and verifying your tagging structure and data collection to make sure you are capturing all the data you think you are. Make sure that all pages are tagged and that custom tags fire properly.
  • Ensuring that all tag parameter variables are accounted for, even if you have no data for a particular parameter
  • Ignoring currency formatting on e-commerce data that is passed in your tracking code
  • Testing all other JavaScript on your site. Any JavaScript errors that occur on a page before your analytics tag will prevent that tag from being executed.

7. Avoid Information Overload

Some organizations go a bit crazy when collecting web data. For example, I’ve seen a client set up a traffic variable that collects an internal search term and then combines it with the page where they went on the site. Yet no report was being used with this information (nor should it have been). Enabling all the parameters you have available can increase the overhead on your analytics tool, and can sometimes cause you to hit limits on the amount of data that can be processed. If any data that you are collecting (other than out-of-the-box) data does not serve a purpose in relating to your KPIs (business goals), then stop collecting it.

8. Set up an Optimization Process

Once you have your analytics program running smoothly, it is time to add an optimization process to it. This involves selecting any aspect of your metrics that can use improvement. For example, an easy win would be to reduce the bounce rate from targeted landing pages, or reducing the exit rate from pages that should lead to a call to action. Longer term, you will want to improve the performance of campaigns to lower your cost per lead or sale, to reduce the fallout rates in your conversion process, or to increase page views or reduce calls to your call center, and so on. Items that can be tested include landing pages, conversion funnel pages, forms, body copy, headlines, offers, colors, graphics, processes and segmentation.

The optimization process starts by implementing a tool that will let you conduct A/B split testing and multivariate testing. Since this is more advanced topic and requires strategic planning and execution to administer properly, you will either want to work with your optimization tool vendor or a company like Edgewater Technology to show you the way. To do this effectively, your organization will want to create a team that merges strategy, technology and creativity together. After you run a given test, analyze your results, make the recommended changes, and test again.

9. Understand How to Measure ROI on Activities

The end goal on any phase of testing is to increase your ROI for that cycle. But, how do you measure that? It helps to understand the ROI formula. Basically, it is the gain from an investment minus the cost of the investment, divided by the cost of the investment. Suppose for example, you have a baseline of an average of 10,000 orders per month from 434,000 visitors. That is a conversion rate of 2.30%. If your average revenue per sale is $50, your total revenue would be $500,000 from these visitors. If, through your optimization efforts, you raise the conversion rate to 3.1%, your resulting number of orders would be 13,454, for a revenue total of $672,700, or a difference of $172,700. If it cost your company $50,000 to make these improvements, your ROI would be ($172,000 – $50,000) / $50,000, or 245%. Note that this ROI was based only on the gross revenue, and does not factor in the cost of goods or services sold.

10. Implement an Analytics Roadmap

Just as a builder uses a blueprint to help guide his team, your web analytics program should also use a blueprint. At Edgewater Technology, we call this a “road map”. It is designed to help move your organization from simply collecting web data to building a comprehensive reporting platform that gives you a 360 degree view of your customer. In this road map, some very important questions are answered, including:

  • Where is your analytics program now?
  • Where do you want your analytics program to be?
  • How will you get there?
  • What are the goals of the various stakeholders?
  • What data to they want to see?
  • What data are you not collecting?
  • Is your collected data accurate?
  • Do you need to integrate online data with offline data?
  • What challenges will you face in getting to your goal?
  • What specific tasks does your team need to do to get there?

Once you have a road map, you will be able to break down all the required tasks and determine what level of effort is needed to implement your analytics program.

Summary

By understanding the fundamentals needed to build a strong web analytics platform, you will be able to provide reliable data that supports your company’s business goals and provides you with actionable insights that can be used to optimize all aspects of your web program.

Understanding Multichannel Analytics

While web analytics can give you a pretty accurate picture of how well online buyers respond to online marketing activities, it fails to tell you anything about how your online marketing affects offline purchase behavior and how offline marketing affects online behavior. If you website has a 3% conversion rate, what about the remaining 97% of your visitors? If you send out 50,000 coupons and get a 2% direct response rate, what about the other 98% of those who got the coupons? Is there a way to measure what they do? Enter multichannel analytics.  Multichannel analytics is a process where all marketing channels are analyzed to develop a more complete view of visitor behavior.

The Four Marketing / Purchase Quadrants

While there are four quadrants of multichannel analytics as outlined in the figure on the right, this post will discuss the two online/offline combinations shown in red. I will briefly explain some of the issues regarding multichannel analytics, some methods of tagging offline marketing and offline purchases, and show you some of the benefits.

The biggest problem with tying in offline efforts or offline conversions is lack of a common point between the two. You have two different databases, one of online data and one of offline data. Unless you have the equivalent of a primary key, you cannot join the two data sets together. Imagine a customer walking into your store or calling your order link and giving you their unique visitor cookie. That would make it fairly easy to tie in their online behavior to their offline purchase. You would be able to track what brought them to your website and what they did before coming to your store.  Unfortunately, in the real world we cannot tie these efforts together, so we need to develop solutions. Solutions for both of the red quadrants will be discussed as they relate to the multichannel analytics integration process, as shown in the following figure:

Tracking Offline Marketing to Online Purchases

There are two solutions to tracking your offline marketing efforts. The first solution is to use vanity URLs in your offline marketing efforts. For example, if you go to DellRadio.com, you will be redirected to a dell.com URL that has some tracking code. In the URL string, you will see a parameter titled “cid”, which is used by SiteCatalyst as a campaign ID. Thus, any purchases from visits to DellRadio.com will be credited to their radio campaign.

You can do the same thing with all of your offline efforts. Put vanity URLs on your newspaper or magazine ads, in your mailers and coupons, on billboards and other forms of display advertisements. Use specific vanity URLs in your radio and TV ads, and simply have your IT department do a “301 redirect” that converts these vanity URLs into coded mainstream URLs that your analytic tool can process.

The second solution to the offline marketing effort is to promote the use of tracking codes in your offline media such as infomercials. Someone watching the infomercial can either call the phone number or order online. If they enter the promo code on the website, you will know that the order was the result of the TV ad. However, what this will not tell you is the percentage of those who came to the site from the infomercial but did NOT buy. If you simply want to allocate revenue to an offline marketing effort, a promotion code will work well with any offline media that drives traffic to your main URL. Within your analytic package, you would tag the code entry as an event, and then look at the revenue that is associated with each event (specific code for each offline activity).

Tracking Online Marketing to Offline Purchases

Now that you have a way to track how your offline efforts work to get visitors to your website, how do you measure what they do when they don’t order online?

Capture Visitor Intent

If your business is both online and retail (physical store), you can measure intent to come to the store by tracking results of your store locator and directions links. By setting these as goals, you can then see what searches were done by visitors who have expressed intent to come to your store. To help capture the buyer while he or she is in the buying mood, some stores like Barnes and Nobles offer the ability to enter a zip code to see if a book of interest is available at a local store. If so, the customer can reserve it online and go pick it up right away. If you can offer this type of service, you need to tag this event so it can capture what brought the customer to the website, and be able to tie in the physical purchase (offline) to the online marketing that resulted in the purchase.

Generate Campaign-Based Coupons for Offline Purchases

It is also possible to have your website generate a unique coupon ID that can be for the particular product that was searched.  By creating an ID that represents marketing segmentation (campaign type, campaign source, media placement, keywords, and so on), you can store this information in both your analytics package and your store database. If you use a campaign translation file for your analytics platform, you will want to include the same campaign ID as a prefix to your coupon. The same coupon concept also applies to service businesses such as insurance, reservations, home and professional service businesses, etc…, where you give the prospective customer a coupon ID that they can use to get a discount. If your business takes orders or inquiries over the phone, you could have your site coded to include the coupon code next to the phone number on all pages. By tracking the redemption of these coupons, you can compute a click-to-store conversion rate, and factor in offline revenue that was attributed to specific online marketing campaigns. This will give you a higher ROI and perhaps provide justification for more web-related investment.

Implement Phone Number- Based Tracking

Unique tracking phone numbers can also be used to measure the impact of your online marketing efforts to offline purchases. A service like Voicestar provides these tools. You can place trackable phone numbers on your site, or use services like “Click to Call” and “Form to Phone” options. Their system has an API that lets you get data right out to your analytics tool and dashboard. Tracking phone calls is very important, as it is human nature to still want to talk to someone on the phone before making a purchase decision. When using a phone tracking service, or even if you have a block of your own phone numbers to use, it is important to not have the phone numbers as a part of the static content. The phone numbers need to be integrated with an algorithm that can associate the phone number with a particular campaign.  To further tie in the visitor to the phone number, a cookie should also be set that relates to the tracking source. Thus, if the visitor leaves the site, and comes back at a later time, the initial campaign that brought him or her to the site will still receive credit for the sale.

The biggest drawback to this type of campaign tracking is that depending on what level of detail you want for your marketing segmentation, you can end up needing dozens or hundreds of phone numbers. This can possibly become expensive and difficult to manage. Instead, you can create a 3 or 4 digit “extension” that is tied to a web-related order number, and when someone calls the number, the phone operator asks for the extension. This has no incremental cost to implement.

Another phone tracking service is offered by Mongoose Metrics. Their service integrates with most web analytics tools to create an automated URL postback after each call is made.  You can perform the same type of analysis, ecommerce conversion and segmentation that you would from any other page to be analyzed. You can see instantly how well your online marketing activities are generating online revenue.

There are many ways to implement phone-based tracking, and they all require integrating your site code with your analytics platform and your backend system.

Utilize Site Surveys to Understand Buying Behavior

Another way to gauge consumer intent is to use online site exit surveys. Companies like iPerceptions, ForSee and others can provide you with surveys that your site visitors can take regarding their online experience. You can ask about the likelihood of them making a purchase offline, and how much their online experience would influence their buying decision. On your online order forms and lead forms, you can also ask the question, “How did you hear about us?” in the form of a drop-down select or radio buttons. Include your offline marketing methods as choices. If the online traffic source is “direct entry”, then you can assign credit for the sale to the way the customer said they heard about your site.

Assign Values to Online Leads

If your business model is to let visitors fill out a form to be contacted by an agent or representative, there are a couple of different ways to tie success (revenue) to a campaign. Some analytic packages let you assign a dollar value to goal conversion pages, such as filling out a request for information form, a pre-application, or other form of customer contact. This dollar value is based on two factors – the average close rate of online leads, and the average dollar value of each deal. For example, if your company closes 15% of all of its leads, and the average deal is worth $500, then the value of each lead is $75 (15% of $500). Thus, your web analytics package can compare that value to the cost associated with generating the lead, and the nature of actions that lead up to it (pages visited, items downloaded, actions taken, and so on). If your analytics tool is set up to give credit to the first campaign touch point (PPC campaign, banner ad, referral site, etc…), you can still assign credit for the lead to the original campaign, even if the visitor does not convert until a later date.

The drawback with this method is that you are dealing with averages as far as the value of a lead. With average lead values, you cannot measure if a particular campaign brings in a higher-value customer than does another campaign. You can, however, get an average picture of how effective your online campaigns are right within your web analytics tool, without having to import any external data. For many organizations, this will provide much more insight than they are already getting about their offline purchases. It does require fine tuning the value you are using as the average lead value, based on your close rates and average dollar value of a new customer.

Track Campaign IDs with Lead Form Submissions

An alternative to this is to create an offline method of tracking online campaigns when a form is submitted. Your campaign code that you use in your web analytics package can be stored in a cookie and submitted as a part of your lead form. If all these leads are entered into a database, the campaign code can also be entered, and later receive credit for an eventual sale. The exact dollar value of the deal can then also be assigned to the campaign, just like for an eCommerce site. The integration of the online and offline data would then need to be done.

Reaping the Benefits of Multichannel Integration

So far, I have touched on some of the ways to “tag” offline marketing activities so they can be read by your web analytics program, and how to tag offline behavior that is due to your online marketing efforts. However, to put it all together requires access to all the data, both online and offline, plus an integration plan that combines strategy, technology, business logic, web analytics data, BI data, implementation, analytics and other disciplines to provide the desired results. One of the benefits of a multichannel analytics integration is that you will be able to obtain actionable insights, such as these (some are industry-specific):

  • Enhanced ROI – Once you are able to assign additional offline revenue to your online marketing efforts and online revenue to your offline marketing efforts, you will see a higher ROI, enabling you to justify additional spending on both your online marketing and other web efforts, such as site testing and optimization.
  • Retail Merchandising Decisions – If your business is retail, your online data can be mined to see what items tend to be purchased together, enabling your retail operation to group these same items together for in-store customers.
  • Upsell Opportunities – If your offline customers tend to respond to particular upsell opportunities when they call in or get called back, you can use this information to target similar online customers or visitors, based on data that can be stored in tracking cookies.
  • Re-marketing Intelligence – If you know what online customers come back to your site to buy later, you can use this knowledge to market similar products or services to your in-house mailing or phone list.
  • Additional Retail Outlets – If you see a significant request for retail outlets in areas that you are not currently serving, you can have the data you need to consider expanding your physical presence.
  • New Promotional Activities – If you know that your online visitors express an interest in finding a store based on looking at particular products that they want right away or that tend to be expensive to ship,  you can create geo-targeted online campaigns that are designed to get more buyers to your store. This can also work well for seasonal or event-driven items (snowstorm, hurricanes, extended deep freeze, etc…), where the need for a product is now, not 7 to 10 days from now. By tracking these click-to-store visitors, you will be able to measure the success of these campaigns.

Hopefully, this post will give you some insight into how multichannel analytics works, some of its challenges, and how it can benefit your organization.