Data is way too expensive!

Who needs data? Not me!

Every senior manager will tell you that it’s better to make decisions based on your gut feeling. We, ourselves, became so accustomed to not having data when we needed it that we took this alternative approach: making decisions based on our instincts.

So why have our sales dropped by 65% in the past six months? Read to the end we will give away a nice prize for your comments

Cutting costs

In an effort to cut back on costs, Lior at Lior LTD online shop decided to scale down the cost of data which had reached over €90K per month, including employee wages, hardware and software. The management set a goal of reducing these costs by 50% by the end of the year, already reaching a 30% cut on Q1.

The BI manager had to decide who to keep and who should be sent home. "We need to review each employee and each of the services we use," he explained to his team, sorry to be the bearer of bad news. He then sent four team members home out of his 14 staff and decided to close some of the services, including an API to collect and update package status along with the most expensive API. Next, he turned to his tracking and tasked himself with reducing events tracking by 40%, and with that, he hoped he would be able to reach the company's goal.

Changes were applied, people left and the team went into firefighting mode. A nice email was sent around the company, entitled: "Please state the priority of your data request". To help employees understand, the email explained that from now on, everyone's requests would need to be queued and would be addressed based on priority and data team availability. Five employees didn’t like this change and resigned, so the data team was left with just five people running the whole show…

"WOW! Now that's a great saving!" said the management, as they congratulated the BI manager.

Saving money on data will bite you back!

But two months after these changes were implemented, the CFO jumped into the management meeting with some very bad news: "Over the past few weeks, our sales have been decreasing and I don't know why – we're still spending the same on acquisition."

The CEO asked the CFO to go and figure out what the problem was, so he arranged a meeting with the heads of Logistics, BI, UX, Marketing and Product to find out what was going wrong. The BI team received a bunch of URGENT requests for data, but with just five people maintaining all the daily data processing, they weren't able to do much.

In preparation for the meeting, each team collected its own data manually:

• Marketing - Collected data from Facebook, Google and other partners with whom they run campaigns, and this data showed that spending was stable, CPA (cost per acquisition) was also stable and that new user sessions were almost at the same level. On the other hand, usage of the CRM tool had reduced by 60%, since package updates were no longer being sent.
• Product - This team collected data from the third party tool, and it showed that session levels were decreasing. They had carried out two changes during this period: removing the package delivery status as it was sunset and changing the product buying button.
• Logistics - Came to the meeting with the shipping log which showed that the time between order placed and shipping was still the same. The packaging had not been changed.
• UX - Explained that they were waiting for data from the BI team to allow them to interview users in a bid to understand what wasn't working. They apologized for not having full visibility, given that they only had the third party tracking tool. Their knowledge was also hampered by the fact that many events were no longer being tracked and e-privacy laws were blocking them from obtaining user consent to share data and better understand users.
• BI - The head of BI explained that since his team was so small and he had a backlog of 342 tasks, he needed to prioritize work based on putting out the biggest fires. Essentially, if something wasn't causing a fire, he and his team couldn't deal with it. He also mentioned that he was searching for more employees and freemium tools to replace the ones they paid for.

What do we do next?

Does this sound familiar to you? Is this something you have experienced first-hand? Would you like to share how you resolved these issues? 

Or perhaps you haven't experienced these problems, but you think you can solve them. If it were you, what would you do?

In two weeks, I'll let you know how we approached the problem, and between the best comments we will give away €10 Amazon gift card -- Limited to 3 gift cards

Article first published in https://www.taleaboutdata.com