Time to kill the “unique”?

Q: When is a unique visitor not a unique visitor ? 

A: Most of the time.

Which as my mother would put it is a bit like saying that you are a bit pregnant.

Unique visitors to websites remains a metric that is commonly used to understand the flow of audience and the success of online content particularly for brand advertising and when ecommerce is secondary or irrelevant to objectives.

These days however a unique visitor metric is not truly unique to an individual in the way that it used to be, so that once again we are measuring things that don’t properly reflect reality. 

If you see a great poster that drives you to a website you may well act immediately, whip out your mobile phone and access it straight away.  That’s one unique visit.  However, it may be that you find that the site requires more involvement that you can manage on the go as it were, so you wait till you get back to your desk and go to the site on your laptop to have a good browse of its content.  That’s unique visit number two.  But just then your boss calls you in to a meeting, so don’t get to really spend time on this very good site until you get home in the evening and go to it on your ipad.  Unique number three.

Three unique visits.  One person.  And no simple way of knowing or agreed industry factor for whether the three visits are one person, or two or three (unless you have registered with and used Chrome in which case Google know).  Now all those visits are very welcome (its not Royston Vasey after all) but it would be really useful to know them a bit better.

Does it matter ?  Aside from the fact that it is an appalling mis-use of language, it matters because we are measuring and predicting success in a way that does not match up with the return on investment.  Planning of a repeat reach based campaign on line may well be based on unique visitors (in the way that it was last year) while the reality of the ratio of unique visitors to how many people you’re really getting as paying customers has changed fundamentally simply after the “tablet Christmas” we have just had.

The reality of uniques to real individuals will depend massively on the type of content and the type of person. A 35 year old ABC1 man with a smart phone, ipad and laptop is not going to behave in the same way of course as a 65 year old C2 housewife who logs on to Facebook on a pc once a week to catch up with her grandchildren. The entertainment category will not generate the same behaviour as snack food.  How much it matters will depend on the kpi of the campaign (if you’re selling a DVD,  so long as enough people buy it, it may not matter how many uniques your campaign attracted, but if you’re trying to change consideration in a category then the tracking of that awareness versus your planned campaign may be change significantly).

There are ways through the data of course.  Yet unique visitors are not what they were or even what they ever claimed to be.  Let’s just get rid of the terminology before we start making too many decisions based on a dodgy metric. At MediaCom we’re changing how we make visitors accountable, but surely it is not beyond the wit and wisdom of our industry to develop a better general measure to drive industry growth in a sustained way.

  • Colin Stone

    Well yes Sue, I quite agree. We seem to be back to the old ‘ optimum frequency’ debate – is 4+ cover likely to be more effective than 1-3+? Even if you get your kpi’s in shape and set your roi targets sensibly, could still be 100% wrong. Funny old game!

  • Richard Foan

    Sue is right to insist that our industry professionals are clear on the metrics used for trading. It is also important that, as with any other channel, digital metrics, once understood, are produced reliably from systems that we trust the configuration of. 

    Registration is a case in point, as Sue says, with Google Chrome, information is available (within Google) for engagement from devices attributed to those accounts. Where I differ with Sue is that the account does not always equal a user. It may equal a household or a team at work and it may not cover all the use by those people of course. In other words, without clarity and transparency of the basis on which the data is derived, planning and buying and selling is open to interpretation. 

    So, what to do? First, understand and use the metrics that our industry has built together – e.g. those at http://www.jicwebs.org. Second, agree up front with your trading partner which ones they will use. Third, insist on a stamp of trust being applied to the systems involved. UKOM for example have based their recent tender in part on the fact that the ABC stamp of trust will be applied to the process used to derive analytics data. In this case, using the ‘unique browser’ (note NOT ‘Unique User’) metric that JICWEBS has endorsed enable a solid foundation to be built for the hyrid data that UKOM (and others) will use in future. 

    Just like circulation in print as the underlying basis for panel based audience data, Unique Browser is not the solution; but it is the only mandatory transparent and universal building block on which to manage and deliver useful audience data in our increasingly fast moving, fragmented, micro trading world which is so full of opportunity for those who make the time to learn.

    Richard Foan, Group Executive Director, Communication & Innovation, ABC

  • Andy Vassilev

    This may take a long and very academic debate, but I will try to be as short as possible.

    In the whole 10 years I have dealt with advertising, which according to some is not that much, I have never seen a metric system that actually represents an individual that consumes a certain medium. Print can account to number of sold copies – but how many people read one copy or how many people bought it just to display it in the waiting room, or even how many media planners have bought it just to have a hard copy of the ad? TV and radio – same thing applies – the number of people in front of the telly or the radio is never known – it is approximated by outdated panels and “hit this button and that button” actions.In fact digital has probably been the only medium that has never claimed that UU/UV are real people. Those have always been defined as machines. In fact I believe that the correlation discussed in the above article exists only in the minds of ill prepared media planners.

  • Most discussed/rated

  • Tags

  • Authors

  • Archives

    February 2013
    M T W T F S S
    « Jan   Mar »