Web Analytics: 5 Tips to Avoid Client Overload

It seems like a paradox – clients that are willing to pay for reporting and analytics and measurement, but unwilling to trawl through the painstakingly prepared report. You end up having to give a presentation but as you are going through the numbers, you see glazed eyes. In the end, your metrics get ignored, recommendations are not taken up, and another month goes by until your next report.

If this sounds familiar to you, here are five tips that in my experience have helped with approaching clients with sound advice.

Drop the Jargon

More than once I get the refrain from clients and partners alike that they were not ‘technical’ enough for web analytics. There seemed to be this impression that measuring web traffic and audience behaviour as analogous to programming an application in an object-oriented language. Although you do need to be somewhat technical to set up and tag your sites, this is untrue.

I believe this stems from the fact that web analytics is replete with jargon – ‘Bounce Rate’, ‘Conversion Funnel’, ‘Correlation’, ‘Visits’, ‘Unique Visits’, ‘Unique Pageviews’, ‘Pageviews’, etc. The list goes on and on.

Break this down into simple language in order to explain these details. For example, instead of “We achieved a 40 percent bounce rate”, say “40 percent of all visitors continued their journey to the rest of the site”.

Avoid Data Vomit

Something that I think we have all been guilty of at one point or another is the data vomit. More often than not, a lot of template reports that got through to the client (or the boss) is just a glorified screen capture, copy and paste PowerPoint that says little about how well (or badly) a site or campaign is doing, and what actions can be taken.

Instead, focus on what is more useful – interpretation and actual analysis. Granted, it is far easier to start digging into the tools and pulling out data but it is not what people want. They want actionable data, and you hold the key to the actionable part. The data is still supremely important, but do not be afraid to show only the fruits of that labour without showing the effort.

Be Selective

Tying in with the above point, remember that the client’s time is precious. Chances are, your analytics will be squeezed in between two other lines in a meeting agenda, and you have been allotted no more than ‘n’ minutes to make your point.

Now, with that in mind, be very selective about the information that you show the client. Granted there will be some clients that will want to drill down to the tiniest detail, but more often than not, they will want to know a few key things – how their property is doing, and what they can do to improve it.

Keep these two in mind when picking the data and analysis to back up your assumptions, brush up on effective ways to structure an analytics dashboard (there are tons of examples online) and clients will appreciate it.

Pretty Pictures

Pictures paint a thousand words, and in this case, pictures save you from spewing out a thousand numbers. Instead of a wall of numbers, include pertinent charts to illustrate what you are trying to show, and try to make those charts communicate that one point.

However, just blindly charting everything is not enough. Too often, one gets carried away and tries to condense a billion dimensions and numbers into one Super Chart. This in my personal opinion takes away from what pictures try to achieve in the first place, which is to illustrate and visualise one or few salient points utilising solid underlying data (which is what an analyst strives to achieve in the first place). Also, pictures of kittens and babies rarely hurt – depending on the audience of course!

Avoid Over-Analysis

A trap that you can fall into is over-analysis. It’s easy to know how to fall into this – when you have a hundred rows of data on your Excel which relies on 10 or more linked formulas to derive a metric, you should watch out. Don’t do this unless you either have a really good reason to do so, or you are relying on an algorithm that has been proven to be accurate and repeatable (e.g., regression formulas or CAGRs).

The reason for this is manifold. Firstly, your analysis should be based on accurate formulation and data, and unless you are a computer yourself, this opens your data up to inaccuracies that can be picked apart. Secondly, consistency is key in reporting cycles, and if it turns out that your formulas are in error, it will wreak havoc on your previous data, and even worse it will wreak havoc on the client’s trust in your underlying methods. Thirdly, it is much easier to explain to the client and easier for the client to trust industry standard tools and algorithms than it is for you to defend the Excel formula that you have just keyed in as an afterthought.


Now of course, these tips are just a starting point, and as with everything, context is everything. Such things depend on your client’s culture, their personality, and also on the property or business need that you are reporting on so one will need to tailor the approach accordingly – and certain points might not hold true depending on the situation. I do hope that these five tips provide a good starting point to thinking about how you can better engage with the client.

Related reading

Big Data & Travel
Flat design modern vector illustration concept of website analytics search information.