The path from data collection to revenue generation is costly and the price continues to increase. We spoke about ways to improve efficiency and get to the insights with Randa Minkarah, co-founder and COO of Transform, a platform designed to optimize revenue by applying machine learning to vast stores of internal revenue and marketing data.
What is the problem you are solving?
The competitor is an internal build — hiring a team of data scientists.
Many companies build over time, build per item. They may have three different financial systems, whether it’s their paid or their organic success in social. What you have is a lot of final data yet you may not see what’s going on in the store.
Is there a proper method for taking
data from silos into one actionable portal?
An alternative is to create a platform where you map the data with a customer to determine what data is needed. We sit down with a new customer and ask for all their relevant data sources. We gain access and ingest directly from the sources into one platform. We automate the ingestion going forward. From there we transform the data into insights. It’s a moving solution and sometimes there is duplicative financial work that can be eliminated.
What factors determine how quickly this can happen?
The volume of data, the variety sources and the velocity of the data, which means how much data they are generating every day. Is it one million records per minute vs. 50 million records per minute is velocity. We can set up the data pulls for continuous ingestion.
Manual adjustment of using spreadsheets stops because we can deliver on a day-to-day basis — really visual insights. Up and down the organization we will deliver the right answers.
Are you selling as many sweaters as your competitor? What did I sell yesterday? What do I need?
On marketing spend, don’t put your TV in this particular market. One store in one market said your market is too big and your impact is too small. This marketing channel is not right for you.
This required all the marketing data to be rejected. They’re doing it online and likely having an omni-channnel approach. The faster we get to results, the better. It’s important on the retail and banking side.
Do you save the “old” data?
Yes, we store historical data in case it is needed/useful.