Vous êtes confronté à des indicateurs de performance produit inattendus. Comment pouvez-vous tirer parti de l’analytique des données pour combler le fossé ?

The biggest problem I find is that people react to data metrics without understanding the cause of the disturbance. One case I came across was launching a spirits brand in a major NY beer distributor who had just expanded into the spirits realm. We had put brand managers into the field and I used my contacts to set up new accounts and sent in dozens of new orders. At the eand of the first month we showed fewer sales than I had generated alone never mind their 50 sales persons and my brand ambassadors. It turns out their techies had not yet set up the brand on the computer systems losing tens or thousands of dollars in launch slaes.

Technology has enabled us to track almost everything, but knowing what data to track is more essential in the modern days. This might mean to not just create a report on "how website is doing", but rather before you ever track or look at data, understand what you are trying to understand. That seems silly, but nail down to the exact aspect which's causing the low performance. For a product page, this might mean looking at Functional links, design, copy, product packaging. ONE by ONE. That's the key. Okay, I get it - It can get pretty overwhelming. But instead of looking at the whole forest to identify where the fruits are being stolen, look at each tree to see if something is off before moving on to another.

If we see any unexpected trends in our product performance metrics we should stop and dive to find the answers to these question “Why does this happen?” Of course we should be able to predict the outcomes of out actions but sometimes this could happen. If we see something like this we first try to find daily, weekly, monthly, yearly data, both data for our product and data for our core competitors because something they did could affecr our outcomes also. If you find an answer in the data than try to find to out what brought this changes, a new campaign, a new rule, a bad execution, a wrong strategy or tactic. You have to search around every aspect that can influence your outcomes.

* Analysing the product trends on a weekly basis is a must for every product person. * If you can't catch up with lots of metrics, (typical for infrastructure and platform products) you can build an alert mechanism on some deviation scenarios * Trends analysis needs to happen on both input and outcome metrics, otherwise we cannot understand the story

Performance analysis always starts with a clear replicable performance workload. Based on observed problems create a repeatable workload with all needed data, operations and concurrent loads/use frequencies modelled in. Once this is in place, ensure the automated test scrips have few knobs like number of users or intensity of load or even mix of operations. With this establish the baseline before starting any analysis. Then start to look at where the problem is. Starting with characterising performance as load is gradually increased and see where the breakpoint is. And analyse system data, internal system statistics, and logs and such. As bottlenecks identified l, fix as needed and iterate. Until problem fixed.

If you found out a main reason for the unexpected performance and outcoumes you should alsı look at all your target groups, segments, to find out which groups or segments drive these outcomes. So you will be able to get in contact with these groups to find new insights you can rebuild your strategy if necessary.

Interview your customers. It's a game changer. Don't just make it all about the best aspects, your product is great, and a good feedback makes it more valuable. Ask them about which feature they wish you had. Ask them which feature they wish can be out of the product. Take notes of all the points of friction, as you do more interviews - if you spot a pattern, working on removing that friction. But also take a note of what their favorite feature is , which aspect they love the most, what is the most frequent user case. As a creator of the product, the aspect which you think is "most useful" for them can be different from what the customers think. That's okay. Use this to optimize the marketing.

Customer feedback is vital for any business. It offers direct insight into how customers perceive your products or services, highlighting areas of strength and those needing improvement. By actively listening to and analyzing customer feedback, companies can enhance customer satisfaction, build loyalty, and make informed decisions that drive growth and innovation.

Numbers are results, but change is driven by humans, so if you see any umexpected results while analyzing your data find out who drives these changes and than go out to answer your customers what has changed and why so you can analyze their behaviours better. Ask proactively for feedback, ask if anything changed with their needs, if they found a new choice that match their needs better and if there is any need to change something on your side

If there is sonething unexpected in your perfomance metrics some competitor could have shaked the market and drive it in another way as you expected. It is not enough to look only at your data changing over time to understand the whole picture, you must look at the data of your competitors also to understand the market and behaviours of customers better. Why did they change their reaction towards us, our product or services, does any of our competitors something better than we? Don’t overthink and get paralyzed with data, just go and ask your customers and collect their answers.

This part is hard but with current possibilities of prompt engineering and access to different indexes, a data driven PM can also make a regular analysis. That should be more exploratory analysis, rather than having a deterministic role in your product related plans.

It is important to have data and dashboards to analyze trends and behaviors, however, this data is useless if it is not actionable. As a decision maker, you must connect the dots, materialize them and put them into action. Do not make the mistake of losing focus, your efforts must always be aligned with the business objectives. do not panic if you see in figures the bad results, on the contrary, lean on them to see the nuts and bolts that need to be tightened. Align expectations with teams and always be willing to improve the user experience because it is a cycle of understanding what is happening, improving and acting on the improvement.