Failures Are Your Friend

Failures can be funny.  They can be also be educational.  Harvard Business Review (HBR) has been running a series on business failures along with a special failure issue.  The HBR message is don’t ignore failures – that is where you learn some of the best lessons but only if you analyze the failure. Harvard has some good business ideas but failblog has more day to day lessons. Hint, riding your skateboard off the roof onto a cactus patch may sound like a good idea at the time but…

Web analytics is great for telling you if an online effort had results.  Campaign X generated Y of orders.  Non-revenue based sites can track a conversion event to measure success.  I wrote in a previous article about the need to identify success points.  Knowing if a particular campaign is working is the first step in optimizing your digital marketing efforts.  Congratulations if you made it that far.  You’ve gone far enough to be aware of the cactus.

Are you looking deep enough at why the ones that didn’t work failed?  Was it the skateboard, roof, jump, or cactus that was the problem?  Instead of going immediately to an alternate selection of a bike off a tree into a bed of nails analyze the first failure.

Advanced techniques drill into the stages in-between to look for failure points.  A common method is a scenario report.  These follow a sequence of actions to identify the drop off point.

You define the path: roof > skateboard > jump > cactus

The report will show visits per step

Analysis shows that most users who enter Roof drop out of the sequence at Jump.  Oh – the problem is gravity!  Just convince them that gravity doesn’t exist and the sales of your cactus spine pulling pliers will increase!  Ok, the analogy is getting stretched but you get the idea.  Overcome the objection at the drop off point by targeting your message at that point.  Maybe it’s a simple interface issue, many visitors couldn’t find the jump function.

But what if the problem isn’t at the drop off point?  Maybe your campaign didn’t fail, it just wasn’t successful enough.  After all there were 9% that made it to the cactus (1 got lost on the way down).  Ask any ecommerce site owner and there are many places they would be ecstatic to have a 9% conversion rate.  Scenario reports can point to the drop off point but you may need to dig further to find the root cause for true improvements.

Analyze what is different between the success segment and non-success.  It may be visitors from a specific campaign or search keyword.  Go outside the scenario and look at the entire path from entry to conversion.  Maybe you are paying PPC for a keyword that gets you traffic but no cactus jumpers.  Maybe the conversion point offer is too specific to the successful jumpers.

Since 9% conversion isn’t that bad you could increase overall traffic.  Measure the methods that brought the most visitors.  If you can increase the 100 to 200 it might increase the successes from 9 to 18.  Still the same success rate but you have doubled cactus landings.  Use traffic source to identify the best source of visitors and increase efforts in that channel.

Analyze the traffic source and drill down into referrer.  If your largest referrer is i-love-cactus-spines.com and you already have 90% of the normal traffic they get putting more effort into that source will not benefit.  Sometimes it is the source that is failing that has the best opportunity for growth.  More failure lessons!  Are there opportunities in the sources with low traffic to increase those?

At the same time you look at improving alternative sources keep track of the conversion rate.  Attracting more of the wrong type of traffic won’t help.  If you increase the traffic from the cactus club site do conversions match the increase?  Keep testing so you know what changes lead to more conversions.

• Identify successes and failures using success/conversion points.
• Use scenario and other path based methods to identify drop off points.
• Look at the drop off point for problems but don’t be surprised if the failure is not at that point.
• Broaden your analysis to look in other places in the path from customer acquisition to the call to action.
• Segment to find areas that affect the outcome and try variations on the segments.
• Keep testing to see what segment change activities work and which don’t.  Look – another failure that teaches us something!

What’s your favorite failure lesson? Share your thoughts with us below or any of our social channels (Facebook, Twitter or in our User Forums)