Numbers do not lie, but they do not always tell you the whole truth either. This is a reality that is very familiar for marketers, who are always pulling out various reports from different sources, only to have the data saying different and conflicting things.

When you have a set of reports – some created by the same person, others pulled by different people – it may seem like they are reporting the exact same concept, but the data is not agreeing. Let us take the example of something as simple as daily revenue.

At first glance we have two very similar reports, but it is also clear that these reports are not the same. They simply look different when you put them side by side with each other. This is not a deal breaker, but it can slow down your process a great deal.

Even worse, the reports do not match. In this case, the daily revenues in the two different reports are showing us two distinct numbers. But there are many reasons why this can happen:

– One of the reports is only about new customers, while the other is for all customers (I came across such a situation yesterday, and I only realized it because someone asked me a random question.)

– One report is targeting gross sales, while the other is including refunds that were processed over the past day.

– One of them is from an outside vendor and the other is an internal report, and the definitions regarding what they are trying to pull are a little different. Sometimes these differences are intentional as companies want to make their numbers look good in various circumstances.

– One of the reports is about cash billings, while the other is based on the GAAP rules.

– The numbers from yesterday change based on when they are pulled – those who are familiar with Google Analytics and Adwords know exactly what I mean.

This is a list that could go on for many pages. Regardless of how much your role or organization is data-centric, it is frustrating when two of your reports do not match. And in a typical situation, if we have two reports that are not matching, there is a good chance that plenty of the other reports do not match either. Everyone wants to be a more quantitative marketer. But these reports that are at odds with each other can seem like they do more harm than good. But do not despair. It is possible to create an analytics program that is providing real value, in quick time, with as little back-and-forth as possible.

Here are six ways to begin:

Step 1: Define Reports Clearly

Start by getting clarity about what a report is meant to say. This is a little easier said than done. But you can ask yourself: If the data does not roll up to a defined KPI, does it matter? When you have defined the purpose of a report, label it prominently in the header or the footer. You do not need to create 40 different footnotes in a report, but you should not assume that people know what a report is because they have been looking at it for a long time either. Find the right balance between precise definitions and having too many annotations.

Step 2: Validate Across Departments

When you have defined your reports, share them across various departments to make sure everyone views them as being the “truth.” This requires some collaboration between the IT, analytics and marketing departments. In some companies, that may be three different people, while other companies may only have one person doing all of these jobs. When all the parties involved in the conversation have signed off, you can validate the reports with your stakeholders. Make sure that the people who are running these reports are on the same page, and that the people who read them also know what they are looking at.

Step 3: Audit Reporting Tools

Business often have far too many reporting systems that they use. I know of some marketers that are using GA, Looker, VWO and Salesforce. Each department likes to get numbers from the system they like to use. Sometimes you have different people within a department who are using different systems because it is what they know. The reports are added to Excel and Google Docs through the network or on a person’s laptop. It is important to determine the reports you need and the software you use to generate them. Standardize this solution across the company and streamline your stack to ensure you do not have a needless number of versions.

Step 4: Designate Owners

Now that you have everyone on board, figure out who owns the metrics. This is tricky, because it is fraught with organizational, technological and political problems. But even though it is not easy, it is a critical step. Successful analytics programs are centralized. I work with many companies that have various reporting tools in place or different levels of implementation and none of these companies has a true owner of report, analytics or the data’s integrity. There are many people involved in the projects, so it is not about having less resources, but it is about using resources in the right way. In my opinion, the owner must be on the marketing or analytics team.

Step 5: Get Rid of Bias

This is going hand-in-hand with the steps we outlined above, but it deserves a separate explanation. Sometimes there are disparate versions of the truth because people have a desire to get their numbers looking as good as possible. For example, if we talk about attribution of orders back to media, some may use a first click vs last click attribution in order to report the numbers. Others may include the view-throughs too. If you layer on the way to map offline media through online orders, things can get even more complicated. These are issues that get even worse when the TV people decide to report to different people as compared to the online folks. You may even get the SEM person reporting to a different part of your organization than the person who is running your Facebook ads.

Defining and validating these reports and having an owner can mitigate some of these issues. But you must make cultural changes. While employees must remain accountable, it is important to ensure that your data collection and reporting is not forcing people to manipulate the numbers to further their own agendas.

Step 6: Outsource

Marketing data and analytics are getting more complicated all the time. There is nothing wrong with having some core competencies in-house, while you outsource the rest. There are so many analytics and attribution experts, which means you will have plenty of options. You can hire an end-to-end service provider to get and report all of the numbers. Or you can bring in a strategic tech partner to help develop your analytics and attribution models. Some platforms streamline a centralized and consistent way of reporting, and they do it in a media-agnostic (or employee-agnostic) way, which removes many roadblocks from the reporting process.

Caveat:

If you decide to outsource, you still have to own and define the metrics internally. And your team is going to have a better understanding of the business than a third party, which means that you must have internal accountability for how the data reports are interpreted and what steps are taken after the reports are seen. But the ownership decision is a little less politically-charged when you have an external company taking charge of most of the reporting.

At the end of the day, it is all about taking strides towards becoming a more data-driven marketer. But I you do not have good data integrity, you are not much better off than those who ignore the data altogether.

Numbers are never the whole truth. But when you interpret them in a holistic way, they can become the basis for analysis. And this is going to lead you towards making more informed decisions and getting better results.

About The Author

Leave a Reply

Your email address will not be published.