Last week, my company hosted its client summit in Boston, MA. It’s an annual event that brings clients together with our team from all over the globe to better understand, inspire, and embrace the digital channel. During the summit, our search team hosted a workshop titled “Getting More Out of Search.” This focused on ways to go beyond the basics and extract more value out of your search campaigns. Most of the time, developing an industry workshop is nerve-racking due to the caliber of requirements and expectations. To create a good workshop you have to: present something of value, say something distinctly new, make it entertaining, and do it all in about 45 minutes. Yeah, good luck with that.
However, this year I found the session pretty easy to put together. While discussing various ideas, we kept coming back to the drum we’ve been beating for quite some time: measure everything. Immediately you’re saying, “That’s not new” or “Just what I needed, more search data.” At first glance you would be correct, but we took a fresh approach to measurement, rooted in data, that effectively allows us to better fund and optimize customer experience.
What You’re Not Measuring
Even sophisticated search marketers are giving the last click too much credit. The final click is not where the measurement story should end. The issue is that marketers are not adding enough weight to other experiences. People may use search as the entry point, but how, when, and where they convert is a whole other story. The average conversion rate is only 2 percent. That means that 98 percent of your site traffic is not doing what you want them to do…or are they? Online conversions aren’t guaranteed just because you created a search campaign. There exists, in fact, a variety of ways to convert that we simply don’t take into account (at least not in a meaningful, measurable way). We should look at all potential conversion paths, assign weight to each, and devise an intelligent way to track them. This will help to illustrate the true value of generic keywords and to stop us from focusing solely on that tiny 2 percent of your converting traffic. So let’s take a look at what you’re not measuring.
Store locator or store visit: Although you searched for “skinny jeans,” you’re not convinced the jeans are as skinny as they look on the model. You fire up the store locator page and head to the store downtown. The jeans fit great! So you buy them, however we don’t count this as a search conversion. There are two easy remedies: coupons and e-mails. Providing some incentive would allow you to capture this converted user and see her search history. If you don’t offer these options, you should at least measure visits to your locator pages, see what Zip codes are most frequently searched, and check store sales data to see if there is any increase.
Latent conversions: Don’t discount your consumer just because they decided to buy outside of your pre-defined conversion window. Aside from stretching out your conversion window, there really isn’t a remedy to this. You can try and survey customers and see how long it took them to buy, but response rate would likely be low. I would simply put this one into a bucket that you know converts, but assign a lower weight to it since you can’t easily track it.
Switched devices: I may start my experience on a mobile device, but convert through my desktop or by picking up the phone. It could be that you have a great mobile experience, but I want to ask some questions to a real person. Or I could have difficulty filling out a form on a small handheld device. Either way, you end up losing me. This is another difficult event to connect, but something that deserves some weight and consideration.
Didn’t convert/bad experience: In my opinion, this is the most important one to understand. What are you doing about the 98 percent of people who do not convert on your site? Maybe its price, maybe a bad site experience, or something simple as browsing that prevents on-site conversion. Regardless of the reason, this is a significant behavior and something that you should not easily ignore. Evaluate bounce rates, keyword segment performance, duration on the site, abandonment rate, and other metrics you deem relevant. Only then will you get a sense of where the bad experience is coming from and can address it.
One of the other reasons this is so crucial is to improve the budgeting process. As you evaluate your 2011 budget, don’t underestimate your search campaigns just because you’re only giving credit to brand searches. Once you start to weigh and measure these additional experiences, you’ll notice important segments of available demand that you’re missing.
Optimizing the Experience
The most unmet need in this exercise is optimizing all experiences, not just search. Let’s look at this fictitious online campaign example:
- You spend $500.
- You get X clicks.
- Two of those clicks convert.
How can you effectively optimize your site based on two people? How do we even know that they have anything in common aside from the fact that they both converted on your site?
The worst part of this scenario is you’re not optimizing the other experiences – you have no incentive to. However, if you better understood what sales were driven to the store, how people who opt in have a higher LTV, or looked at bounce rates you could assess those channels and optimize them. They could be simple fixes like creating more content or providing better incentives, but if you don’t measure them, you can’t optimize them.
There are, of course, more conversion-related experiences that we didn’t discuss. The matter at hand though, is looking beyond the “last click” for areas you should be evaluating. Only then can you piece together a much more complete conversion story, highlighting those experiences that are weak. And only then can you track and analyze to better optimize the experience.
SEO and search marketing are a vital part of any marketing strategy, linking together channels like social media, content marketing and offline advertising.
There is of course a lot of discussion about content and what does and doesn't work online. Is long-form the key? Does short-form content have a role to play? Are there other factors at play?