I don’t consider myself a brand or direct marketer, just a marketer. I emphasize the value of branding, which frustrates direct marketers. I talk about the value of measuring what matters, which prompts brand marketers to email corrections.
Several months ago, a reader asked an interesting question:
We will often report the average time spent on a Web site. This intuitively seems like it should have a value we could wrap into our ROI [return on investment], but… it stands largely on its own. Do you have any thoughts on how we might put an actual value to this? Is it enough to show lift without respect to time and talk about return visits in terms of frequency models? Or is there some way to [drill] down to a fundamental value of what a person-second on your site could be worth? The question generated heated debates among my colleagues. We even got Jim Novo, former VP of marketing at Home Shopping Network, and Roy H. Williams, three-time bestselling author of “The Wizard of Ads” trilogy, to chime in. I hope you’ll find the summary of the debate as valuable as I did.
Brand marketers understand advertising value in terms of brand affinity, recall, intent to purchase, and so forth. So that’s what they measure, making a final connection between those metrics and revenues. Direct marketers measure audience composition, audience size, audience response, and final conversions. Most attempts to apply one set of metrics to the other discipline fail. Not surprisingly, practitioners on both sides consider their strategy the superior one. Only a few marketers on each side of the fence are willing to look over and (without admitting the grass is greener) ask for some seeds, at least. So let’s attempt some cross-pollination.
Direct marketer Jim Novo said: Traditional advertising has never been judged by the value of the customer; it is judged by the value of the media. The customer is reach and has no individual value. Individual customers are totally exchangeable as long as the reach is the same…. It doesn’t matter what they do or don’t do. If there’s no customer, I’m not sure how you’d ever get to ROI. It’s assumed that from reach comes sales. This is proven using branding metrics, not ROI.
If I were going to try to “straddle” direct and brand metrics, I’d migrate toward evidence of loyalty. You can use frequency, but [it’s] better when combined with recency. Not only have they visited often, but they’re still visiting. Since recency predicts repeat action, you can imply someone who has visited in the past seven days is more loyal than someone who last visited 60 days ago, because she’s more likely to visit again. You can look at the average recency of the visitors created by different campaigns and measure which campaigns generate visitors with the highest average loyalty (recency).
This is exactly how direct/database marketing companies manage customer retention: falling recency equals defecting customer. Falling recency for a brand marketer’s Web site could equal falling loyalty. Loyalty rising or falling is a metric I think branders have a good understanding of.
This makes some sense if you think of it in terms of demographics. If you look at recency by search engine, you usually see dramatic differences. Visitors coming from one engine are more loyal than those coming from another. This generally has to do with the distribution (and thus demographics) of the engine. Each search engine is really like a cable TV channel, with its own demographics. You can further see differences within a search engine by topic, which is similar in concept to the different demographics of shows on a single cable TV channel.
As for time spent on the site, it’s pretty difficult to comment on without understanding the objectives of the site. In every case I’ve seen, longer visits equals higher sales, leads, downloads, etc., because tasks take time to complete. But unless you’re selling time, as with traditional media, I’m not sure time has an economic value to the marketer.
I’ve recently studied data that proves Jim correct about search engine demographics. My recently published report, “Converting Search Engine Traffic,” ranks the top 15 major search engines by best visit-to-order conversion ratio. The report is based on numbers from more than 29 million search engine visits to dozens of e-commerce sites using WebSideStory’s outsourced HitBox services and collected by its StatMarket service. The source and term of a search are critical because the source defines the context of the term. A person conducting a search for “desktop PC” on CNET may be exhibiting a greater intent to purchase than someone doing the same search on MSN.
Time also has economic value to consumers in terms of opportunity costs. They could easily be doing something else rather than staring at your site. Time spent on your Web site could be viewed as the sum of attention people are willing to give you, instead of something else.
Roy H. Williams, the Wizard of Ads, puts it this way:
Time and money are proportionate. The company that receives the largest percentage of a customer’s time is the most likely to get their money.
Average time spent on a site should be measured relative to sites that offer comparable products or services. It is, without question, the most reliable single indicator of customer interest next to actual sales.
There’s no other correlation that can be reasonably made. The variables are far too diverse to compare TV costs to the cost of time spent on site. This may come as a surprise, but I’d much rather have the customer spend 30 seconds on my Web site than 30 seconds watching my TV ad.
Agree? Disagree? Let me know what you think.
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
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
Google Analytics comes with lots of standard reports and settings, but with a little customisation you can extract much more value. One way is ... read more