Search our Site

3 minutes reading time (525 words)

Impossible or incredible? MEDA's rolling baseline methodology


Imagine starting a business with the mandate of having 20,000 customers within the first week. Sounds incredible, right?

Customer bases build over time through advertising, word of mouth, new partnerships, and sometimes expansion into new product lines or geographies. It makes good business sense to start small, test a business model, and improve it as you expand.

MEDA approaches business development the same way.


NicaraguaMEDA staff registering project clients in Nicaragua

Rather than reaching out to 20,000 small businesses within the first month of a project, we often bring clients on over two to three years, typically within the first half of a project. If we’re going to fail, we want to fail small and quickly and continue to improve our approach for existing and future clients.

Even hidden within the most obvious successes, we can find better ways of doing things moving forward.

This rigorous and methodical approach means we need to do Monitoring and Impact Measurement (MIM) differently.

A single start-of-project baseline (‘before’ snapshot), with neatly-timed follow up surveys simply does not make sense.

MEDA applies a Rolling Baseline Approach to capture baseline data from clients as they enter the project, and then follow up annually to see how things may have changed for them through working with us.

How do we keep track of 20,000 clients’ before and after snapshots this way – is our MIM department made up of superheroes?

Well yes, it is, but our rolling baseline methodology is designed to keep things manageable too.

We register clients upon their first interaction with the project by collecting their name, location, nature of their business, and contact information so we can find them later. Then we include them in a cohort of clients that have registered within the same time frame, say, every six months. Every six months, then, we would conduct a baseline from a sample of these registered clients, in most cases ensuring a statistically representative sample by end of project. We go back to this same cohort every year to conduct a follow up survey.

This graphic shows how it works:


Digging a little deeper into the rigor of this approach, we can consider how it could be used as a control group methodology. The idea of using control groups is always met with controversy because of the ethical challenges associated with it.

How can you say ‘no’ to a potential client simply because you need him or her to be included in your control group?

And there are methodological hurdles as well.

For example, choosing a control group from a different geographical area than the one you are working in means your group may not be all that representative of your target clients. The rolling baseline methodology, however, allows us to compare two cohorts’ data as long as the groups are deemed to be sufficiently similar. A particular cohort’s baseline data for date x can be used as a control group for comparing an earlier cohort’s follow-up data for the same date.

MEDA’s business development work aims to mirror any good business model: it is progressive, adaptive, constantly improving, and our baseline methodology just ‘rolls’ with it.

How business support helps emerging-market enterpr...
How to host a world night (or is that Rain?) marke...

By accepting you will be accessing a service provided by a third-party external to