Prioritize Your Optimization Opportunities
Updated · Feb 15, 2005
You use a great Web analytics tool. It delivers great data tied to your site goals and to supporting metrics. Through analytics, usability studies, and surveys you identify several opportunities for improvement.
The next challenge is to prioritize these opportunities, as you’ll likely compete for resources to act on them. You must ensure you focus on your site’s most important elements and get your team to agree on the priorities.
We help clients prioritize changes by creating a list of all their initiatives and opportunities, then monetizing the potential effect of each one. This requires assessing the upside of each site change, assigning a dollar amount to each, and comparing them to one another.
Below, three common opportunities for improvement and ideas on how to monetize their value.
Consider improving your calls to action and overall lead generation process, then assess how much you think you can affect them. If the lead generation’s conversion rate is 20 percent, can you drive it to 22, 25, or 50 percent? If you can, ask questions such as:
- How will increasing the conversion rate affect the number and quality of leads?
- Will I generate more leads of significantly lower quality, or can I keep the lead quality good while increasing the number of leads?
- What’s the lead’s value?
Once you determine a lead’s value and the results you believe you can gain with optimization, you can calculate the upside of the proposed changes.
Also, calculate how quickly the improved site can pay for the optimization. Look at monthly and annual effects. Now, recalculate the potential result of driving more people to your form and improving conversion.
The same basic principles apply to online commerce. Again, assess the outcome of improving the conversion rate of a single step in the purchase path, and tie it to your overall conversion rate and bottom line.
Remember to look at both revenue and profit. You probably don’t want to spend $50,000 optimizing something to increase sales by $200,000 for the year if your profit margin is only 10 percent. In some cases, this can make sense when considering the lifetime value of new customers and so forth.
Again, consider how you can affect the bottom line by tuning the purchase path. Don’t focus on one specific number; look at a range you think you can achieve.
If these changes pay off and the upside is still there, you may go through three, four, or more rounds of optimization to continually drive a specific metric higher. As long as the return outweighs the costs (money, resources, and opportunity cost), it makes sense to keep tuning.
Customer service differs from lead generation and e-commerce quite a bit, but we can still compare its potential to the other options once we monetize the results. Online customer service often focuses on reducing the cost of servicing customers while still satisfying them.
See if you can improve customer service on the site to reduce the number of inbound calls and e-mail messages. It’s typically less expensive to service people online through self-help content.
If you can deflect 1 percent of all call center calls per month, you may save hundreds of thousands of dollars each month, or more. Evaluate the cost of servicing someone on the Web versus through the call center. Look at ways to swing the numbers more to the Web, and see the results even a very small percentage can return.
Compare Apples to Apples
Maybe you want to improve all three opportunities or multiple cases of each one. Now that you’ve monetized the outcome of the changes by estimating the ranges, you can stack them side by side. Using these findings, you can work with your Web team to prioritize changes or to fight for more resources and budget dollars.
Though I can’t share specific ranges of realistic improvements for these different types here, I’m happy to share realistic estimates offline. Feel free to shoot me an e-mail. I’ll speak at eTail 2005 in Palm Desert, CA, this week, so let me know if you want to talk Web analytics!