Regular readers of this column know that I’ve been talking a lot about how to set up for a serious data-analysis effort — specifically, determining which measurements will have the most impact on business decisions.
Many readers have written to say that the question of what to track is actually standing between them and the start of a data-analysis program. So I’m going to start taking up some (anonymous) readers’ examples and sharing ideas regarding how their specific business situations might benefit from better data collection and insight.
A site developer from a content site reports that the editorial-content folks have a strong interest in tracking usage by page and subject matter to determine what is of greatest interest to the readers. That’s a reasonable start, but this developer wants to know how to make the measurements relate to monetizing the site. Good question — one everyone is asking these days.
I can suggest a few very different approaches, all worthwhile, but depending on the content and purpose of the site, one may make much more sense to generating revenue than another.
If the most likely source of revenue is subscription fees (assuming the content is not readily available elsewhere and has high perceived reader value), then the next step would be to postulate which segment of current visitors might conceivably convert to paid.
If you profile some likely paying customers, you can stop looking at total site traffic and instead look at where those likely buyers spend their time. I’d recommend developing several alternative profiles so that multiple hypotheses can be tested before you risk all on a sudden move to a pay format.
This assumes some sort of registration process so that you know whose behavior you are watching, and it assumes an ability to count far more than eyeballs. You want to look at the pages a given type of visitor spends time on, how time spent and the frequency of return vary by parts of the site, and whether there are repeat patterns of travel through the site.
If subscriptions are unlikely (as they are for most content sites), two other likely revenue streams are ad sales and transaction revenue. Planning for either requires deep audience knowledge — the more specialized or desirable an audience, the greater the revenue potential.
In the case of transaction revenue, pay attention to every action and the patterns of folks who do complete transactions. Also, identify ways that audience segments differ from the larger audience segment that never quite gets around to closing the transaction or that does so too irregularly to generate profit.
In the case of advertising, it’s a little trickier. You need to determine which subsets of your total audience are likely to be of interest to a large-enough share of advertisers. Then you need to research how much advertisers are willing to pay for that audience while you research the behavior of the visitors on your site who match that advertiser-sought profile.
Nobody said it would be easy, eh? But it is straightforward: Understand how you expect to make money, and then segment your audience into those who can make you money and those who cannot. Don’t write off the group that is not paying for itself now, but do think about ways to redesign or restructure the site to better entice those others to become revenue builders for you.
Keep looking for patterns that produce your desired result and those that do not, and use that data — plus insight — to guide all your product design and marketing decisions.
Next week, we’ll look at some other challenges.
Emily Ma, product director of Tencent’s advertising platform products department, was a keynote speaker at ClickZ Live Shanghai where she discussed the ... read more
In today's multichannel world how can marketers use data to ensure the experience a customer receives is relevant to them?
The terms that customers type into your site search function can help you to gain an understanding of user behaviour and can be used to optimise ... read more