If I don’t clean my house for a few weeks, I start getting dust. In time, the dust builds into little clumps here and there. (This is normal, right?)
To extend the analogy to customer data, the challenge is the same. Customer data from various sources (surveys, competition entries, app sign-ups, prototype hand-raisers, event attendees, and more!) accumulates in small pools, which over time get older, less relevant and dustier.
This “small” data is often particularly tricky. You might not be able to integrate it easily into your main dataset because it is gathered from some specific set of interactions, or it’s relatively too small to do anything meaningful with.
Most marketers talk a good talk about using data more effectively, especially targeting those niche audiences with the stuff they want us to talk to them about.
But we also know that when the spirit is willing, the flesh can be weak. If you need to hit targets, you may not have the time to squeeze small fry datasets, even if rich growth opportunities lie within.
Let’s pause and exercise some empathy for those marketers.
It would be easy to ignore a pot of 10,000 records when, say, your total pool of records is 2 million.
IT departments are also a barrier. Effort from their side to integrate dusty data is often the same for small or large pots of the stuff, and so if the size of the data pot is relatively small, they’ll simply tell you politely to ‘jog on’… not enough value to be gained.
For iota-ML, size truly doesn’t matter. We don’t complain if datasets are too small, because we have ways and means to do ‘data alchemy’, i.e. to turn your dusty data into a goldmine of new and exciting targeting strategies and growth opportunities.
Our clients often come to us for two solutions:
OPTION ONE: Take small, awkward, “Kim & Aggie”-level dusty datasets and have the big brains of the iota-ML data science team insert it directly into your system, to be used directly against your customer base in clever and meaningful ways.
OPTION TWO: Apply the iota-ML predictive process. Cross-reference this dusty data with your total pool and spot look-alike customer audiences, then craft the right message to talk to this totally new segment. Voila, you have ways to create impactful, targeted marketing strategies leveraging data that would be otherwise ignored!
What does option two look like in real life? Say you work for an airline and you have the responses of 10,000 customers who filled out a customer satisfaction form, complaining about their recent travel experience. iota-ML could take those customers and find lookalike audiences in your huge transactional dataset. For instance, customers who travelled at the same time / were on the same plane / spent time at the same airport / bought from the same in-flight menu, but didn’t complain.
Suddenly, you have an opportunity to proactively engage these ‘silent detractors’, perhaps offering the unique experience of a targeted apology coupled with a personalised promotion (e.g. “sorry there was no hot food on your recent flight, here’s 10% off your next return flight valid for the next 12 months”)!
This is just one simple example of the power of hyper-personalised, hyper-relevant marketing powered by Machine Learning.
This is something ever more important in this recessionary age, and this is something iota-ML can help with.
It’s quicker, easier, and costs less than you may think. To find out more, reach out to email@example.com.