There’s no doubt that a hot topic for 2022 is in-housing. But what’s the right solution for you?
Historically, it has been typical for marketing effectiveness, in particular Marketing Mix Optimization (MMO) programs, to be outsourced every year or so. The output from which is often a backward-looking view of what worked and what didn’t.
However, there is limited value in this approach when it comes to future strategy. As more firms recognize this, we are seeing a distinct shift towards ‘always on’ MMO. Which in turn is driving the trend towards in-housing. But there are many considerations around personnel, technology, culture and approach.
In-housing is not for the faint-hearted, especially for organizations with multiple brands in multiple territories. And there are many variations on a theme to finding the right solution – there is no ‘one size fits all’ approach. But there is a way to structure your thinking.
There are three main models: Partner-led, Hybrid, and Fully in-house.
In-housing’s allure is in the ownership and control this brings. However, the right data science expertise can be difficult to find and attract, which is where a hybrid approach can help. To determine which model and where best to employ partner help, consider the following:
It’s essential to be clear about what you are trying to achieve before determining whether in-housing will deliver value. For example, is this requirement global, local, long term or short term?
Sales or customers?
For any brand that is concerned with customer outcomes, from driving new customers to delivering value from existing customers, privacy, and the associated management of data in partnership approaches, becomes critically important.
Ownership & Control
Fragmented landscapes with multiple providers, from media to data and analytics, can be messy and unwieldy. Bringing it in-house to own the full end to end process facilitates a holistic solution, which can then also be used for other analytics purposes too. Plus every piece of insight can use consistent, native language and be successfully shared across the business.
Scale & Transparency
Scaling up with an outsourced partner can be both difficult and expensive; product and brand complexity can quickly mount up to several hundred if not thousands of models. AI and Machine Learning can help here, but the trade-off is often transparency. Internalising ensures you can see exactly what’s going on and provides a single basis on which to build those models. It also allows for expansion and replication into other products or markets.
Agile & Holistic
This in turn allows for a holistic and agile approach to respond to new challenges. It becomes easier to see which levers have what effect, add new channels, and meet the changing needs of different brands around the world. There is greater opportunity to optimize and speed up insights. Though on the question of speed, there are many considerations about how fast insight is really needed, as real-time is expensive and may not deliver the incremental value the cost demands.
To make all of this happen, strong data, both internal from customer behaviors and external such as media, retailer, competitor and macroeconomic influences, is the foundation to a successful capability. This can be a complex project in itself, but one that is necessary from the outset. Ask if you have the basis of this capability already, or if you need to build it.
Skills and expertise
There will always be a balance between corporate memory and broader expertise. Critical questions include, can you attract and retain the right people? Is what you are asking of them reasonable? Do they know how to embark upon a project of this magnitude? Can they deliver the evolution such a project needs? Do you have longer-term opportunities for them to make sure they stay?
Breadth or depth?
Aiming for breadth and depth at the same time is unlikely to be successful. Whether you choose hybrid or fully in-house, you will need to take it one step at a time and design a program that delivers incremental progress.
Begin with one or the other. For example, focus only media, perhaps because you have a single agency providing the data, and just control for promotions, shopper and instore for now. This way, you can iterate data and people processes, such as increasing frequency, in a step-wise route to depth as you roll out across the business.
Or, choose full automated depth of data, a key driver of scale, but for just one market. Perhaps one that needs the most attention. Here you can replicate the template and path to data integration, having learned from best practices in the first markets.
The route you take will depend upon a multitude of factors. From the platforms and agencies you use, to your business model and markets, and your analytics needs. Either way, the end point is achieving both breadth and depth as you build out.
Are we ready?
Building this kind of capability takes commitment. Are you ready to take on a roadmap that, for complex organizations especially, can stretch over several years? Is there C-level buy-in that will also push collaboration and adoption? Do you have the experience, data maturity and end user adaptability to take it on? Do you even spend enough on marketing to make the investment worthwhile? Can you manage the expectations and needs of numerous stakeholders?
If you think you might be ready, you can delve deeper into this topic in the Ekimetrics whitepaper, Marketing Measurement & Optimization: in-housed, outsourced or hybrid?
Matt Andrew is a data science specialist and MD of Ekimetrics