Igor Souza

An RFV matrix (recency, frequency and valor) can help you not only identify the best customers, but also churn risk and behavior. Different channels, message and incentives should be optimized for each case so you extract and recover the most of your user base. This brings a richier vision than short time span metrics like user engagement, pages per session, avg session time, etc. because you don’t rely on recent user visits and signals.

A mind mapping tool like Miro or a UI tool where you can freely draw simple things like Figma. Which tool is not the most important, but focusing in mapping everything in a clear and concise way.

You should also consider your main objective in your category plan. For example, if you plan to raise the average amount of purchases the users do in one year, you must invest in a category that leverage this purchase behavior in the consumers, like convenience or basic needs products. This way, your business and category goals connect and contribute to themselves.

A working DSP strategy is not one-size-fits-all, even inside your own company. Each supplier, distributor, warehouse, and even some categories of products and sales have different timings, and you must consider that or all your effort will be in vain. Good DSP solutions and consultancies will help you identify and keep track of those needs. Although overwhelming, this detailing of your operation is one of the biggest benefits of DSP. Identifying and addressing these differences are not agile or simple to track, but after you mapped it all you may achieve an agile management, but most importantly, effective and efficient DSP.

I must say, this is not an easy task, neither for small nor big ecommerce. The best advice I can give is to start assuring that your day-to-day “basic” tools are well implemented and that you take a look at their data architecture. Do you trust your GA4 data? Know how to extract useful data from your ecommerce's platform? Is your CRM and your customer data organized and reliable? The reason for that is, when starting a DSP strategy, you most surely will need to extrapolate the data from where you manage your stock and enrich it with more data from these platforms and more. From there, you need to understand the best way to extract and centralize all this information in one place.

Every research must lead you to an answer that is meaningful to your business and moment. Knowing what your business truly needs is a challenge, but researches can lead you to discover that also. You should be open to hear the users and not use closed questions, limiting the answers too much, or filtering too much who your research reach. Start broad, validate and discover hypotheses, then go limiting and being more specific along your unbiased discoveries.