TOP MISTAKES OF NEW DATA AND ANALYTICS LEADERS

Updated: Jan 28


Ninety days goes by fast (if you read The First 90 Days). You are running a sprint. You need to think about things but have little time to do it. If you step into one of the following pitfalls, you may not get out; at best, you will lose a shoe, affecting the rest of your run. Staying with the analogy, running a sprint with just one pump will not help you break any records. You may injure yourself or even get disqualified from the race. (Note to self - this analogy is closer to reality than you think!)

Avoiding these mistakes will help stakeholders that matter perceive you as THE RIGHT PERSON FOR THE ROLE. Your boss may have named you, but it's everyone else you have to convince that he made the right choice.

Do what made you successful before This is the old "sticking to what you know." It got you here, so it is tempting to keep doing it. Because the context and assignment are different, and the job more visible, the same recipe will not apply. People will question your choices and challenge you. You will be offered some help you don't want or need. And if you did not explore all the scenarios, be careful as you may be overlooking critical elements. Picking the right path is a lot eliminating the wrong ones, if not just that. No path chosen will ever be perfect, so you have to be ready to defend your choices, and you may not be as familiar with what the best (or least worst) one implies. Don't fall into the trap of sticking to what worked before.

Act right now We may be talking about a sprint, but don't fall prey to "action imperative." The wrong actions can amount to a lot of rework or opportunity costs. Don't gamble or go the trial-and-error route. Understand all the strengths and weaknesses you need to contend with and identify all the opportunities and threats facing you BEFORE you act. As we mentioned in part 1, you do need to work fast, but to take a tennis analogy now, better to hit an ace than win an exchange that tires you out and shows your opponent weaknesses he can take advantage of later. Take another breath or two and concentrate. It will be worth your while.

Do too much Accept that there are limited budgets and that you should not plan to overstretch your team from the start. If they enjoy their work, they will give you the extra mile when it counts. In football, whether you win 50-12, 25-12 or 18-12, you still win. There will be unknowns, so you have to allow for contingencies. A 30% contingency is advisable: 15% is the best practice, and the other 15% is the extra that people will ask you to do - for no additional budget, of course - people like your boss. Try not to say no to your boss. So do not commit to too much because, despite all the excellent work you are doing, there will be surprises. You can only be ready for this with contingency in your pocket, but you probably don't want to label it that way, FYI. Commit to less, deliver all of it, and give a few extras. This goes hand in hand with the next pitfall.

Set the wrong expectations Of course, you say. But look at all the work you need to do in the first ninety days to even begin setting these expectations. And you could have tripped three times before getting here (this is pitfall #4, by the way)! I have worked in consulting most of my career, and I will say it now even if I seem biased: get outside counsel now. Your business partners do it all the time with the Big Four, with their lawyers, their marketing agencies and their financiers. They do so because they want to turn every rock, do not want to be surprised, and they need to be backed up by experts. In the first ninety days, this kind of advice has the best return for the least investment - maximum leverage. The value is very high. Make sure you see clearly and consider everything BEFORE you begin setting expectations. Don't be a hero. Consulting firms have frameworks to make sure nothing gets excluded from the discussion, and all relevant decisions get taken. How do you sell this to your boss? Data has become too critical not only as a strategic advantage but to survival. Your competition is probably getting the right advice. Your organization cannot afford to keep striking out and not getting on bases (that's a baseball analogy, here).

Give them "the answer." I knew a consultant who taught courses in dimensional modeling for years. He had already helped another company set up a sales data warehouse. When he was offered a Data Engineering Director role in another company, in his mind, he knew what to do. He hired an army of data architects with experience in star schema modeling and Azure/SQLDW because that is the recipe that made him successful in his previous role. This approach is particularly risky before carrying out a proper assessment; if the context is similar, great, but it often is not.

What if he later realizes that the business people in his new organization are very data literate and they do not need nor want the data relationships to be pre-ordained by him and his team. So here he is creating lags and delays for data preparation the business does not need! It may even hurt them and prevent them from doing what they need to do with the data.

Focus on or learn the wrong things The scope of your intervention is more comprehensive than when you were an individual contributor, and therefore there is a lot of learning to do. It is natural to want to spend more time on the things that we have more experience with or that interest us more. After all, if it is an area you know well or are passionate about, at least you have that on your side - no one said a promotion meant you had to suffer (too much). True. But you know where I am going with this. As a manager or director, you have to assume that the area you would naturally be attracted to learning about is likely not be the area that needs the most attention, and you and your team may not know enpugh about it to even know.

To get things done quickly, like 90 days, you need to surround yourself with expertise with an objective view and a have a process to help you sift through everything - and the current team you have cannot help you at the moment. They should be involved, but as a leader, you need to make sure they get gently led in the right direction, and that often means away from their comfort zone and yours.

Identifying only some of the stakeholders Your boss is obvious. But do you need to involve his boss? Why would you? What about colleagues? What about on the business side? For example, when we propose to change course like you probably will, chances are some important stakeholders were there and possibly instrumental in setting up the previous data environment. What happens if you do not involve them and "seek their sound advice"? The sad truth is that you can do all the right things and still suffer. You can build it, and they will not come, or even block you. If you are a career IT person, change management may not come naturally. Make sure you use a framework that forces you to look at all the aspects of what you are trying to embark on and seek guidance, internally or externally.

©2020 by Modern Data Analytics