Web Analytics: Focus on Optimization in 2006
Working on next year's budget? How to discover the largest potential optimization opportunities.
Working on next year's budget? How to discover the largest potential optimization opportunities.
Over the past few months, I’ve written a number of columns on Web site optimization. As a proponent of all things Web-analytics focused, I’ve repeatedly said analytics is a waste of time if you don’t use the information to drive change on your site and to improve overall site performance. Analytics’ ROI (define) is zero if you only look at the data and don’t act on it.
A zero ROI can be a concern when you look at the investment many companies have made in analytics over the past few years. A medium-sized to large organization can easily spend $300,000-$500,000 per year on analytics, including licensing, maintenance, and tag placement. You must invest in both Web analytics and maximizing your site based on analytics data.
Unfortunately, most organizations don’t budget for Web site optimization. As you create your Web budget for next year, total up your Web initiative and Web budgets and create a new line item for site optimization. Plan on dedicating approximately 5 percent of the overall Web budget to this effort. When calculating that 5 percent, make sure you include all Web-related costs (internal and external), such as:
As you add these items up, you may realize 5 percent can become quite big, especially for larger organizations that buy a lot of media to drive traffic to the site. Include media in this calculation because organizations that spend more on the Web channel and media typically have a larger potential upside in optimization. And it will greatly improve the returns on your media campaigns. In addition, newly identified testing and tuning opportunities can return astonishing results when you haven’t looked at testing before.
Allocating even a small percentage of the overall budget to a new line item can raise questions, especially if you ask for more money than last year. But once you explain how this 5 percent will be used, it may prove to be the easiest budget justification you make. The 5 percent often pays for itself easily and quickly and will likely improve the overall Web site return, helping justify every other Web-related budget line item.
If you need to justify this new budget line, pick two or three of your top key performance indicators (KPIs) or key conversions, and look at their performance over the past 12 months. Next, model the financial impact those would have if you improved their performance by a small amount.
For example, a site has lead generation as the KPI. Last year, the company converted 2.38 percent of all site visits to register as leads, returning 2,800 leads per month. With the typical lead closing rate and the average value of the ultimate close (you can use lifetime value here as well), you can determine the lead’s value, in this case $213. The current month value of the leads the site will generate is $596,400 (2,800 leads multiplied by $213 lead value).
Through optimization, you can forecast an improvement of different levels to understand optimization’s potential outcome. The monetary value of incremental percentage changes in the conversion rate:
Though the above model assumes lead quality remains the same, it’s easy to see how improving just one KPI by a small amount can lead to large improvements. If you model this for a number of your KPIs, you can begin to make a case fairly quickly for the 5 percent budget line item. In addition, once the value of testing is understood, it becomes more logical to do it. Once you do this, you and your coworkers will be astounded it took so long to start testing.
Begin by focusing on campaign landing pages (online, pay per click, etc). Although this often can return great initial results and help improve ad buy ROI, don’t stop there. On a typical site, most people don’t come in through a campaign when they come to your site. They may have come through a campaign on a previous visit, but they don’t often convert on that initial visit (depending on the product, offering, or call to action). Many return later through the home page.
I’m more than happy to share examples and stories on how to model, monetize, and forecast opportunities, as well as share our experience with different types of testing methods, including examples with multivariate testing tools such as Offermatica and others. Feel free to email me with additional questions.
As you put those budgets together for next year, look for the largest potential optimization opportunities!