Here is the fact – web analytics is a cost center. Just like finance, HR, and many other support functions, the web analytics team provides a service to its internal customers. No matter how you spin it, it is very hard to effectively tie revenue growth or business goal accomplishment to web analytics. Sure, you can weasel your way into getting a web analyst head approved for a major project. And when there are cost cuts, that’s the first thing you lose. So how does one justify an ongoing commitment to web analytics?
There is a need for culture change. But that change doesn’t come easily unless you are the CEO, CMO, or COO of the company. You will need to prove the value of web analytics in little steps. To be clear, I am not talking about getting web analytics implemented at your company. To get ideas on how to justify that, here is a good resource. What I am talking about is taking the basic installation of web analytics with perhaps a couple of web analysts and moving it to a game-changing, always in-demand team that actually makes a visible contribution to the company’s bottom line.
How Web Analytics Helps
Let’s recap at a high level what web analytics can do for your internal customers. Web analytics can tell them what happened, when it happened, and even how it happened. What it cannot do is tell them why it happened. Instead of why, we can come up with fairly accurate suppositions about the root cause for a certain change in metrics. After that, to validate the assumptions, there needs to be an investigation outside of analytics or an A/B test across these theories.
Using Web Analytics at a Small vs a Large Business
There are several things in common in using web analytics across different types of businesses. For instance, the tags have to be on the pages no matter what, someone has to run/maintain reports or do ad hoc analysis, and someone has to make sure the tags didn’t magically fall off the web pages.
However, there are a few things distinctly different between small and large businesses when it comes to web analytics. For instance, people at small businesses are typically not dedicated to web analytics. It is one of their half dozen other responsibilities. So if you are in this situation, you need to focus on automation as much as possible. I cover this point in the productivity improvements section below.
The challenge for analysts at a larger organization is not about using analytics but about using analytics effectively. As many analysts as there are at a large organization, ask yourself whether they are doing true analytical work or pulling reporting and emailing them all day long. Because teams typically grow organically and not in a highly organized and thoughtful manner, inevitably communication silos are formed and the web analytics practice as a whole becomes grossly inefficient and ineffective. More on the optimal structure of analytics teams in another post.
So as you read through the ways to improve the state of analytics at your company, carefully consider the approach that best suits your situation.
The Soft Sell
Before pitching any of the ideas below, prep your audience by making (entirely true) statements of the following nature:
- “If we can’t measure it, let’s not build it.”
- “Our understanding of our business will never increase if we continue to work at this level.”
- “We are still struggling with the basics of web analytics whereas other companies our size are now investing in xyz.”
These are just examples, of course. The point is that you want to do a soft pitch before you bring out your big ideas. Get the decision makers interested in what you are saying. Remember, at the end of the day, you are a marketer. You are selling something, and you need to make sure your message is strong and well received. So wait for the right opportunity and make statements that will get people to invariably ask you the question “So what do you think we should do?” Bingo. This is what have been waiting for. Now walk them through a few ideas on how to improve as described below.
Why Invest in Web Analytics
So why indeed should one invest in analytics? Simply put, because without such an investment, a business cannot leverage analytics to its true potential. There is a growth and maturity cycle of analytics in an organization. Without investing in analytics, the analytics team is never able to reach full maturity. The team is stuck doing the same exercise over and over. The days change, the executives change, but the reports essentially say the same things they did before. So what do you do? Here are five reasons to invest in web analytics:
- Improve Productivity
- Maintain Integrity
- Ensure Collaboration
- Foster Learning
- Continue Regular Upkeep
When talking to decision makers, though, don’t bother talking about the last two. Focus on how your decision maker can benefit from the investment. So pick from the top three – productivity, integrity, or collaboration. Let’s dive into each one to understand exactly what I mean.
Analytics can be a funny business. Everybody wants reports delivered to them all day, every day. But if you tracked how many people actually opened the reports and did anything with the information, you would be appalled. OK, so it isn’t possible to wean people off reports or supplement their satisfaction of having access to one. It is, however, possible to reduce the churn and time spent creating these reports by automating them. You will be surprised at how even the most complex reports can be automated to save an analyst’s time. We have seen analysts spend a full four months of entire year doing nothing but running reports. So the case for productivity can be very easy – take the number of hours spent each week creating reports. Multiply by the average hourly rate of an employee. Multiply by 52 and you have yearly savings. Multiply by the total number of analysts wasting their time doing the same thing across the company and pretty soon you could have a robust case for saving the company hundreds of thousands of dollars!
Savings = [Weekly hours spent creating manual reports x hourly rate of employees x 52 x number of analysts]
The case doesn’t stop there. With the cost savings, you also have freed up additional analyst bandwidth. So now you can put that to use in analyzing real problems for your sponsors’ teams. What a win-win!
Here’s another productivity play – I believe in the concept of a hub and spoke team. I can’t imagine any other structure working well for analytics. That is a bigger discussion for another post, but you can sow the seeds now. Every business team owner wants his/her own dedicated analysts. Sometimes it is a power trip and other times they truly believe they can run their team effectively if they had dedicated resources. Play this to your advantage – educate and offload all the mundane reporting tasks to your sponsors’ teams. So you can focus on the harder, more satisfying, thought provoking, data analysis tasks! So the pitch here is to offer to train other analysts to improve productivity – your productivity.
For a website where changes are made frequently, gaps in data or concerns about the validity of data are a way of life. A typical web analytics administrator spends up to 20 hours a month just investigating and fixing issues caused by changes to the site that were not communicated or tested thoroughly. The problem is multiplied many fold if the website uses a ‘rolling release’ approach where changes are pushed out whenever there are enough of them ready. Talk about an analyst’s worst nightmare! There are several levels of solutions to address this problem.
First is the process solution where an administrator could be part of a release cycle process, signing off on releases where the analytics changes have been considered. Then is the tool solution where a second tool could be used to validate the data integrity of the first tool. And finally, there’s the regular automated audit solution where audits can be run on the live site every day and alerts can be sent out when breaks or changes are detected.
All this requires money, which, of course, requires justification. So the formula to justify investing in an analytics integrity check at your company is similar as before. Estimate the hours an analyst could save by not having to deep dive after every release or change. Add it to the number of hours spent fixing and annotating reports. And then the hours saved because of having to triple check everything when integrity checks are not in place. The hour count doesn’t stop there! Include the time a developer then spends on fixing the issue, a QA spends testing the fix, and a release manager spends managing the change! Multiply it by the number of releases in a year and you will soon have a strong case for developing a robust integrity check program for web analytics.
Savings = [hours spent deep diving after every release/change + number of hours spent fixing reports + hours spent triple checking numbers constantly + hours spent by developers to fix issues + hours taken by QA to validate + release management hours] x [number of releases/change each year]
Large organizations often experience productivity losses because of lack of communication. Either that or they are notoriously inefficient because of bottlenecks caused by highly sought after resources. In other words, projects slow down because the right resource isn’t available or projects are implemented incorrectly because the correct resources were not consulted in the development phase.
So if you are going to pitch this one, then you should use the “if we can’t measure it, let’s not build it” statement in a conversation with the decision maker. See how this all fits in?
To ensure collaboration, you need to be appropriately staffed. So make a list of recent projects that did not have the right analytics in place. Of course, choose projects that the sponsor is close to. Then paint a picture of how with additional dedicated resources you will be able to better service the organization’s needs.
The bottom line is this – to leverage web analytics to its maximum potential, an organization needs to make a commitment to nurturing and growing web analytics. Anything less results in overworked, inefficient, mediocre analysts working with poorly implemented web analytics software.