Noise or Music?

Should you focus on website visitors as individuals?

Categories: GDPR & Privacy, Metrics Understanding / Comments: 22

I came across an interesting eConsultancy interview with John Squire of Coremetrics on how he differentiates from other vendors.

I have a lot of respect for Coremetrics as a web analytics vendor that tries to be different. Rather than using the common sales hyperbole of saying “our tool is better because we have 10,000 features more than yours” (no-one cares about features that don’t provide insight, right?), he describes a number of areas that Coremetrics has worked on. I will let you read the interview in full, though I wanted to comment on one particular aspect…

According to Squire, Coremetrics is differentiating its technology by focusing on individual customer data in a multichannel environment. Linus Gregoriadis (eConsultancy interviewer and all-round good guy) makes the comment:

"This focus on website visitors as individuals makes sense for web analytics vendors
because a significant limitation of Google Analytics is the inability
to tie up analytics with data at an individual customer level."

This is perfectly true, Google made the decision not to track individuals in 2005 during the acquisition of Urchin Software Inc, the company behind Google Analytics. At that time, individual visitor tracking was in beta testing and in fact went on to become part of the Urchin Software suite – the server-side web analytics product that is also provided by Google.

I have discussed my take on the issues of privacy and Google before. To summarise, the view at Google (from the very top of the organisation i.e. Larry, Sergey, Eric), was that individual tracking, even anonymously, is a step too far and the web public does not want this. That view remains today, and I am a strong advocate of this.

Leaving aside the issue of privacy, is it valid to track visitors as individuals?

From a marketer’s perspective, tracking individuals sounds great in theory – you understand your customers better right? But if you receive 10,000 visitors per day and have weekly marketing performance meetings, that equals 70,000 data points to discuss? Best practice is to consider longer time frames in order to mitigate against calendar anomalies i.e. weekends v weekdays, holidays, the weather, force majeure etc… So for one month that could be 280,000 data points.

Of course no-one is going to look at each visitor session, so the first thing you do is either:

  • Aggregate i.e. group visitor types together. For example, visits that add-to-cart, visits lasting longer than 60 seconds, visits that viewed the special offers section, returning visitors who have previously purchased etc.)
  • Sample a smaller selection of “representative” visits for analysis and the potential quagmire of what that means. For a discussion of sampling considerations on the Internet, read the excellent sampling review paper of Nigel Bradley (University of Westminster)

Aggregation obviously defeats the point of tracking individuals in the first place.

Sampling, unless done with scientific rigour each time, introduces its own errors. So for me, tracking individuals (even anonymously and with visitor knowledge/consent) is not so relevant for the marketer or web analyst.

What do you think?

Share Button

Comments (most recent first)

  1. @Lukas – thanks for the feedback. My point about integration is that for the vast majority of web sites, the vast majority of visits i.e. 95%+, are completely anonymous to you. For these, integrating with a CRM is not viable.

    Essentially, there are much bigger gains to be had on focusing on improving the quality of engagement with the 95% segment, than the 5% segment. In addition you avoid the two BIG issues of tracking individuals – privacy and accuracy.

  2. I’m not really sure why you are arguing against the trajectory of web analytics integration with business intelligence and CRM systems. Once a customer logs in or responds to an email or survey or requests information – I think the relationship between web analytics and the customer changes.

  3. Justyn: I kinda of agree with you. If a visitor has given you their personal info (e.g. email address), then I see no problem in tracking them – you just need to make sure they are aware of this in your privacy policy. However, if you don’t have personal info, don’t track them at the individual level. I view that as very black and white.

    Apart from protecting their privacy, I just don’t see how tracking individuals is going to dramatically help you optimise your website or its marketing. After all, you still have all the referrer details.

    Megan: Great input. Sounds like you are well on your way to building a unified web measurement platform.

  4. Megan Douglas says:

    Hi Brian,

    It took a lot of convincing to get us where we are, and we have miles to go. In my opinion, the more a company diversifies how it collects and actions data (we have web analytics, partner (ad) analytics, doubleclick and others for marketing, affiliate tracking, and transactional data to name a few), the more work you are creating for yourself with your resources in the wrong places.

    Our company is finally looking to use our analytics first party cookie and its unique identifier to start linking all our data sets together. In my opinion, web analytics is the glue to all online company analytics – there isnt anything web-based you can track via another vendor that you cant also track in analytics if you put the effort into it.

    At least that is true of our business which is a strictly online business with no brick and mortars.

    I realize this is a bit off topic so I apologize, I just wanted to answer Brians question.

  5. Justyn says:

    Great discussion here. My personal view is that their are degrees of privacy. I don’t track every element of a visitor and link it to a profile where I know all kinds of personal data.

    Rather, I will use GA campaign parameters to link to individual email accounts to track “hot” customers and see who is responsive. Linking that data to very general demographic information would be amazing, but that’s where I find myself approaching the fuzzy line of privacy invasion. That’s where I want the customer to volunteer data, not me extract it without their knowledge.

    Back to my original point, privacy isn’t a black and white issue, in my opinion. Certain elements of data can and should be tracked.

  6. Megan Douglas says:

    Thanks for this post. I had been having this argument with a colleague whom I respect. This article helped solidify my opinion that unless you know the user and have a way to contact them (customer), the idea of obsessing about uniques is useless.
    Our company blurs the lines between Customer Analytics and Web Analytics (web analytics is in the Business Intelligence dept which is under the same overall department as CRM). We tie our analytcis data to our transactional warehouse and our CRM system. In that situation, yes, uniques are the name of the game. Otherwise, it would require a ridiculous amount of effort to capture uniques only to be left with significantly inaccurate data (for all the reasons noted above.) I am glad to see this topic getting so much discussion.

  7. Sharique Ahmad says:

    The only reason I would use an individual’s data would be to track any anomalies – missed orders or any other transactional error but that information can also come from an eCommerce vendor’s own system tracking.

    Marketer’s should see segments as a first priority and then individual visits/visitors to test or find any anomalies or exceptions. However, how many marketers would have the time or utility to do such analysis.

  8. Fand adams says:

    Looking for a unique blog visitors is my obsession.. Open the sites review iPhone and Blackberry

  9. r4 says:

    I feel that if we visit a company’s website it is a little like walking into their shop/office – it would be normal for someone to speak to you, no?
    The internet is not public, these sites are owned by people and businesses so I feel we should expect to receive feedback from site owners.

  10. webandrank says:

    All the visitor we want tracking in not counsumers better focus in how we tracking the peoples, because is peoples not data.The intelligence is not individual is collective.
    Just my opinion.

  11. Michael F: The distinction I make between web analytics and customer analytics is a real one. For web analytics, 95%+ of your data is from anonymous visitors – you know nothing concrete about them, other than they came to your site at a certain time.

    Customer analytics is at the other end of the spectrum – you know all the detail they provided in order to become a customer i.e. they are 0% anonymous. You simply cannot merge the two and call it ‘business intelligence’ or similar.

    For example, if Omniture sent me a personalised email because I visited their site, they would be guessing. It would likely contain a sales type request when, if you know my profile, you would ascertain that this the last thing on my mind if visiting the Omniture web site.

    Bjoern: Agreed on your insight. My point is not to ignore individual data if you have it. However, focusing on the 3-5% of data that is not anonymous, means missing the much bigger and more lucrative part of the pie. If your conversion rate (i.e. obtaining contact details of a visitor) is 5%, that means 95% of visitors rejected your attempts to court them. My view is that focusing your efforts on understanding the 95% rejection rate, will reap you far greater rewards than the focusing on the other.

    Shelby: As always, privacy considerations are always on my mind… Respecting end user privacy will *always* generate you more business in the long run.

    Michael B: Same comments from me as to Michael F. I would add that I agree that as soon as a visitor converts the relationship changes and so does the ability to track them. However this post was about tracking web visitors, not just the small percentage that become customers.

    For example, understanding your “top 10% of web customers” from a site that has a typical conversion rate of 5% means you are looking at 0.5% of your visitors. While of course that is an important segment, my view is that focusing on the other 95% will reap *far* greater rewards. In my view its a crime(!) to focus on 0.5% of your web visitor base when there are such opportunities available. Its why most web sites are rubbish

    Sami: I agree, scary stuff. Essentially this comes about from the acquisition of Double Click – which surprised me at the time. I can almost see Eric grimacing at the thought and know this will not sit comfortable with him. Nothing is set in stone, so I guess we just have to wait and see.

  12. “the view at Google (from the very top of the organisation i.e. Larry, Sergey, Eric), was that individual tracking, even anonymously, is a step too far and the web public does not want this. That view remains today, and I am a strong advocate of this.”

    What about this piece of news then:

    “On Mar. 11, Google (GOOG) said it will begin to offer ads using what is known as behavioral targeting, which tailors ads to people’s interests and online behavior. “We’re looking to make ads even more interesting,” says Brad Bender, a Google product management director.”

    To me that sounds like individual tracking, no?

  13. Michael Bain says:

    I disagree. Customer level data is not about analyzing 70,000 separate data points or just sampling. I think those are straw man’s arguments.

    It’s about matching customers with their requests. It’s about customer insights and marketing and being able to link web analytics data to other systems. The goal of the web analyst in this is to help ensure that “important” website events are tracked.

    I’m not really sure why you are arguing against the trajectory of web analytics integration with business intelligence and CRM systems. Once a customer logs in or responds to an email or survey or requests information – I think the relationship between web analytics and the customer changes.

    And even the customer expectation changes. The customer is now in a fulfillment relationship where they *expect* something specifically based on their responses or actions, be it through a back end system or whatever. How do we measure this unless we have customer level data? And it’s more than tracking “thank you” pages.

    Seems to me that companies are beginning to ask the question: How do we understand our top customers experience? Or positive and negative response customers? How can this be done effectively without details at the customer level? Who are your top #1-10 customers and site users and what are they doing this week?

  14. Great post Brian. I completely agree.

    Bjoern’s example of customizing user experience by opting in is very exciting. I just wish Omniture would figure out how to do something simple like offer 1st party cookies out of the box before they try to be anything and everything for their clients.

    I can see the point that Michael makes as well. You still need an account login at some point to trigger those individual communications based on the web analytics data, though. Many sites that use web analytics are not e-commerce sites.

    I also am not sure you *can* put aside privacy when talking about this issue. I obviously understand why you did for this post, but it’s just too important a topic not to address, in my opinion.

  15. vinay says:

    depends upon the business. If you are an e commerce retailer who focuses on particular range of age people (zappos for eg), it might make sense.
    Else it might be too overwhelming

  16. Bjoern says:

    Thank you for your post, Brian.

    I agree with the privacy concerns completely, and I am not an advocate of individual user tracking, because of this.

    However, I think there are scenarios, where individual tracking becomes very meaningful – and that is, when you plugin your web analytics data into other applications.

    Let’s imagine an e-commerce site that sells mobile phones and is able to track the pages viewed by each customer individually.

    If I can now segment and track the logged in users, and monitor which of them puts a mobile phone in his basket, but does not complete the conversion, then I can send him/her an automated email offering a few days later, offering the phone at a slightly better price, by plugging the analytics data into my email marketing solution.

    so, in my view: Looking at individual web analytics data doesn’t make sense. Plugging individual web analytics data into other marketing channels could produce meaningful results.

  17. I agree… good post 🙂


  18. Brian,

    Interesting comments. however, I’m not sure I agree with you (and Stephane).

    In your comment to Stephane you mention Customer Analytics. This is the direction web analytics is moving in, albeit at a rather slow pace in most cases. So I don’t really understand the distinction between the two. I see them as merging or overlapping.

    Yes, individual tracking is imperfect and there are many data integration issues. However, I know of a B2B company using web analytics in combination with real time behavioural targeting to personalise user experience. Using an opt-in data collection mechanism they are able to create a full profile of potential clients.

    They trigger site surveys, target ads on media sites and test offers on their site, all based on the client profile. They even send personalised communication based on actions taken on the site (and I don’t mean a simple name merge at the top of an email but complete customisation of the communication content).

    The company is Omniture and they use this programme both to increase their lead generation and show case their technology integration.

    You are right that manually analysing the behaviour of individuals is ineffective. The key is automation. We will reach a point in the future where integrating the different tools will become simpler and more common amongst middle of the road companies.
    At that point in time web analytics tools would absolutely have to provide individual level data to feed other systems.

    Sure, these solutions won’t be 100% fail proof but the added value will outweigh the disadvantages.

    Thanks for getting the debate going.

    AEP Convert

  19. I salute Coremetrics (and others) objectives to track “unique people” instead of “unique cookies” in the hope to get more precise and detailed information, which, in most cases, is worth a lot. But this might well be a lost cause… the inherent technologies of the web don’t allow that, unless you can authenticate every single visitor (through username/password)… As soon as there is “multichannel” and “web analytics” in the same phrase I get suspicious… how is data from call center, back-end core systems and other touch points integrated to give a “single view of customers”? We’re heading in the realm of business intelligence, and despite the marketing fluff, I don’t see any current web analytics vendor really playing in this space.

    I like your argument that aggregating & sampling is the way to go. After all, web analytics is closer to statistics: with a population, segments, sampling/aggregating… and a margin of error, confidence and significance… then it is to core systems where each piece of data should be 100% accurate and traceable.

    One of the thing playing against Google Analytics is the fact that experienced analysts (those who have a statistical background, come from the offline world or have lots of experience on the web) want to audit, validate and diagnose, from time to time, the raw data and the lowest level possible.

    (longer take on my blog at )


    • BClifton says:

      Stephane: I agree – Customer Analytics, with its sophisticated data mining techniques, is a very different world to Web Analytics where the vast majority of data is from anonymous visits and where the term uniques is meaningless.

      As you say, I don’t feel these can be merged into one “common currency” (I detest that phrase) of measurement. Rather, they should coexist – with insights taken from the overlaps in the data points, when present.

      BTW, if you want the raw GA data for auditing purposes etc. – run Urchin alongside Google Analytics. Assuming a best practice implementation of both tools, the numbers can get very close indeed. Look out for a post on this in the near future 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *

Anti-spam question (required):

This site uses Akismet to reduce spam. Learn how your comment data is processed.

© Brian Clifton 2019
Best practice privacy statement