Data, data, data… It sometimes seems like it’s all marketers ever speak about. But how can we actually go about deriving useful insights from the wealth of data that’s available to us, and what sort of things should we be looking for in the first place?
On 18th June, we hosted Sepi Pouryahya, Chief Data Science Officer at Fospha, and Adam Broitman, Senior Expert, Digital Marketing at McKinsey for our New York attribution breakfast at The Nomad.
Sepi began by introducing Fospha, a marketing attribution technology provider that builds MMMs (mixed marketing models) to power CDPs (customer data platforms). Fospha works with a large number of companies across a range of verticals – financial services, ecommerce, retail – to solve their marketing attribution challenges.
Content produced in collaboration with Fospha.
All marketers want to save money, save time, and increase the efficacy of what they’re doing
In fact, marketing usually boils down to two main questions:
- How can I reduce my CPA?
- Where should I spend my digital marketing budget?
Accurate measurement is essential to answering these questions. If you can’t work out what’s actually working, you don’t know what to prioritize going forward and what to avoid.
Simple, right? Well, not so much. We all know that businesses tend to operate in silos – there’s the marketing team, CRM team, sales team, and others all trying to take credit for any one conversion. In fact, there was one company that Sepi worked with where, if you counted up all the “conversions” that each team took credit for, it was 5 times the number of actual conversions.
But it’s not their fault – they’re simply using the tools at their disposal, which shows them a touchpoint and then a subsequent conversion. However, marketing measurement is about so much more than simply identifying the one single touchpoint before a customer converts. How often do people buy a product the first time that they come across it?
So, how can we actually go about measuring our marketing efforts?
A recent BCG (Boston Consulting Group) report cited three main steps:
- Connect the data
- Automate integrations
- Have some actionable measurement
This all sounds easy in principle, but there’s more to these three steps than meets the eye
Sepi continued by explaining how Fospha have approached this.
Stitching and collecting the customer data
In recent years, especially given the overwhelming focus on data, marketers sometimes forget that they’re actually trying to talk to a customer. Marketing isn’t about lifting metrics, it’s about reaching customers on an individual, personal level. This means that Fospha approaches everything from a customer-first approach: looking at how the customer behaves, what their customer journeys are like, and working backward from there.
The key to accurate measurement lies in stitching together all your marketing touchpoints. It can be difficult to apportion a $100 conversion across multiple different interactions, especially when you do this in-house – every team wants to take as much credit as possible, for obvious reasons. Therefore, having someone who can independently stitch together your data is the best way of measuring your marketing efforts.
Turning these measurements into actionable recommendations
As mentioned before, metrics matter little if you don’t actually know what to do with them. That’s why Fospha focuses on providing actionable recommendations based on three critical components: customer data, online data, and offline data (all of which are brought into a central data store, a CDP). Once it’s processed, you need to work out key identifiers and make recommendations based on those.
Start small and grow
Don’t try and do everything at once. Instead, find something useful and valuable, and look to change this before trying anything else.
Sepi was then joined by Adam for a Q&A session where they touched upon the fact that many marketers currently treat machine learning and data analytics simply as a buzzword, a “line in a powerpoint slide”, rather than actually deriving any useful insights from them. However, the tide is turning, and the best marketers are starting to use machine learning to give them a truly 360-degree perspective of any one customer.
And it’s not just about customer data: context data can reveal useful trends, too. For example, whilst the price of oil isn’t necessarily linked to your company’s product, it can be representative of overall economic and political trends which have an impact on your sales. Machine learning algorithms might be able to find a link between the two that you hadn’t previously considered!
In all, it was a wonderful event and thanks to both Sepi and Adam for their fascinating insights. We can’t wait for the next one!