Customer Profiling and Outlet Segmentation for African FMCG Brands

What is customer profiling and segmentation?

Customer profiling is collecting the facts about each outlet. Segmentation is sorting outlets into classes that get different treatment. For a brand selling through shops, the customer is the outlet.

Profiling is the data. Segmentation is the grouping built from it. Sales teams in Lagos, Nairobi or Accra often call the second step outlet classification. Profiling without a class is just a survey. Classes without profiling are a guess.

Why does classing outlets by eye fail?

Size and potential are barely related in general trade. Roughly 80% of FMCG retail spend in Africa runs through general trade outlets (GeoPoll), so the eye test misleads most of your calls.

Take the largest provision store on the road. It looks like Class A. But it gives your category 10% of its shelf and sells mostly other brands. For you it is a small outlet.

Now take a kiosk that is already busy with your product. Its shelf is full and the owner will not add lines. A quieter duka two streets away sells less today, yet it could double with weekly calls and a better range. Gut feel sends your best rep to the first and forgets the second.

Floor area tells you little about throughput in kiosks in Lagos, dukas in Nairobi or spaza shops in Soweto. Profiling adds the missing half: what an outlet could buy from you, not only what it buys now.

Where do you get an outlet list when shops have no addresses?

You build it, through an outlet census, also called field KYC. No purchasable outlet database exists for most African markets.

You cannot segment what you have not mapped. So profiling here starts one step earlier. Reps walk the territory, geo-tag each shop, photograph the shop front and record the format and the owner's name.

Treat the census as the base of your route-to-market plan, not a one-off project. Outlets open, close and move. A census that is never refreshed is fiction within a year. The same census also profiles outlets that do not buy from you yet, which shows your coverage gaps and the new shops worth a first visit.

What are the four types of customer segmentation?

The four classic types are geographic, demographic, behavioural and psychographic. The same four are often called the types of customer profiling, because a full profile holds data of each kind.

They were built for shoppers, but each has a direct version for outlets.

Behavioural data often follows RFM: recency (when the outlet last ordered), frequency (how often) and monetary value (how much it spends). It comes from order records, so it needs no extra capture on the visit.

The last row is where potential lives. Potential is what the outlet could buy if you won a fair share of its category. Geography and format tell you which outlets to cover. Behaviour and potential tell you how hard to work each one. For how formats differ, see general trade vs modern trade.

Which data points should reps capture in every outlet?

Capture the eight fields below on the rep app during the visit. A profile written from memory back at the depot is a profile built on intuition.

Each field must change a decision. If nobody will act on it, drop it. Reps abandon forms that feel like paperwork, and half-filled profiles are worse than none.

How to profile and segment outlets in six steps

Decide what the classes will drive, capture the data, score sales and potential apart, set class lines, give the scoring an owner, then test and rescore.

1. Decide what the classes will drive. Write down the decisions each class changes: call frequency, which rep covers the outlet, schemes, range and targets. If a class changes none of these, you do not need it.
2. Choose the data to capture. Pick the fields above that predict sales in your category, and add order history from your own records and your distributors. Keep the form short enough to finish on a normal call. A short form every rep completes beats a long one nobody finishes.
3. Score current sales and potential separately. Current sales come from order data. For potential, rate catchment, category sales, space and owner attitude: one point for low, two for medium, three for high. Four factors give a score between 4 and 12. Keeping the two scores apart stops a busy outlet with no headroom from looking like a growth outlet.
4. Set the class lines from rep capacity. Outlets with high sales today are Class A, whatever their headroom, because that volume needs protecting. Outlets with high potential and low or middling sales today are growth outlets, Class B, and so are steady mid-range outlets. Outlets that are low on one score and no better than middling on the other are Class C. Then check the workload. If reps cannot call every Class A outlet as often as planned, narrow Class A from the top third to the top quarter until they can.
5. Give the classification an owner. Reps collect the data, but they should not set the class, since a higher class often brings more support to their own outlets. A sales operations or trade marketing analyst runs the scoring, and area managers review any outlet that moves.
6. Test, then reclassify on a fixed rhythm. Try the classes in one territory for a quarter and compare productive calls and sales with a territory that kept the old classes. After roll-out, rescore every quarter, and sooner after a refit, a new competitor next door or a change of owner.

To place each outlet, rank current sales against similar outlets in the same territory: the top third is high and the bottom third is low. Call potential high at 10 to 12 points, medium at 7 to 9 and low at 6 or less. The grid gives the class.

Example: one duka scored and placed

Take a mid-size duka on a busy road in Nairobi, near a primary school and a matatu stage. Today it buys little from you: a few fast-moving lines, ordered once a month through the distributor. On size and past orders alone it would sit in Class C. Its profile says otherwise:

• Catchment: high. Steady footfall from parents, students and commuters.
• Category sales: high. It sells a wide range in your category, mostly other brands.
• Space: medium. The owner could free a shelf if your range sold.
• Owner attitude: high. The owner works the counter and asked the rep about a new pack.

That adds up to 11 out of 12. Low sales and high potential make it a growth outlet, Class B. It now gets a call every two weeks instead of once a month, an entry scheme on two core lines and a target set from its potential. Once its sales reach the top third of similar outlets in its territory, it moves up to Class A.

How does each class change visits, schemes, range and targets?

Segments set call frequency first, then money, range and targets. Use the table as a starting point and adjust it to your category and your rep capacity.

Visits. Rep time is your scarcest resource, so the class should feed straight into the journey plan, which many African teams also call the call cycle. Class C outlets often do fine with a phone call between visits. Measure the result on strike rate, meaning productive calls divided by total calls, and on lines per call. Transport costs in Africa run near eight times the world average, so every wasted call is expensive twice over.

Schemes. Untargeted trade promotion spend gets pocketed. The outlet takes the discount, volume does not move, and the scheme repeats next quarter because nobody can prove it failed. Segment-level tracking closes that loop. A loyalty scheme for Class A should never look like a trial scheme for a growth outlet.

New launches. Most new products fail on placement, not on product. Send a launch first to outlets whose profile fits it: the right price band, the right shoppers and room on the shelf. Then the first 500 outlets are chosen rather than accepted.

Targets. Split the territory target by class. A Class A outlet near its ceiling cannot grow as fast as a growth outlet, so each class needs its own goal.

Which 10% of your outlets drive half your sales?

Sales are concentrated, and that is why segmentation pays. In Lagos, a detergent stocked in 100,000 outlets can take half its sales from just 10,000 of them.

Most brands cannot name those 10,000. Coverage reports show outlets billed, not outlets that matter, so effort spreads evenly across a market that is anything but even.

BeatRoute's own research finds 30 to 40% of outlets underserved and 10 to 20% overserved, with roughly 6% territory-level leakage. Classes are how you move calls from the second group to the first. That is a coverage and route optimisation problem before it is a software problem.

How do outlet classes differ by industry?

The method stays the same across industries. The data that predicts sales changes.

• FMCG and confectionery: many small outlets that order often. Footfall, shelf and cooler space, and order frequency drive the class, and call frequency is the main lever.
• Pharma: chemists and doctors are profiled separately. For a chemist, the prescriptions written nearby and whether it stocks your molecule decide the class.
• Building materials: fewer, larger hardware dealers. Credit terms, counter staff influence and project business decide the class, and calls are longer and less frequent.
• Lubricants and auto parts: workshops and spare parts shops sit in one network. A mechanic's advice can matter as much as the shelf, so profile the workshops too.

Customer profile, persona and ideal customer profile

These three terms do different jobs. A profile is facts, a persona is a typical character, and an ideal customer profile is a target.

• Customer profile: the facts about one real outlet, such as this provision store on this street. It changes as the outlet changes.
• Customer persona: a made-up character that stands for a group, such as the owner-run kiosk that buys on credit. Personas help you write schemes and sales pitches.
• Ideal customer profile (ICP): the kind of outlet most worth winning. For a brand, it is the outlet type and size you most want to add to your coverage.

Profiles feed your segments, each segment can be summed up as a persona, and the ICP tells reps which new shops to go after.

What makes a segmentation model go stale here?

Four forces age an African outlet database fast, and each has a fix you can run this quarter.

First, price bands shift with currency. The naira fell 40.9% in 2024, and a shelf price recorded before that move now describes a different outlet. Re-capture price and pack mix, and never freeze a class on an old price band.

Second, people leave. Frontline rep churn runs 25 to 35% a year, and each departure walks off with outlet relationships nobody wrote down. A structured profile is the only version of that knowledge the company keeps.

Third, duplicates inflate everything. 'Alex Super Store' and 'Alex Sup St' become two outlets, and your numeric distribution quietly lies. Their order history also splits, so neither record looks like a big customer. Fourth, stock crosses territory lines daily through open markets like Idumota, so sell-out at an outlet may never match what your books say it received.

Four more mistakes to avoid: classing on size alone, and classing once and never again. Do not run so many classes that two of them get the same treatment. Do not let reps set their own class.

Why should your distributor care about your segmentation?

A shared outlet list ends the argument about whose numbers are right, and better targeting protects the distributor's own margin.

Distributors are business owners you court, not command. Their worries run in order: claims that take 8 to 12 weeks to settle, dumping into their territory, and margin under currency pressure. Segmentation touches all three.

Order history for outlets that buy through distributors comes from the distributor. These are secondary sales, and they only help if each distributor invoice carries the same outlet ID as your rep app. When both sides see the same profiled outlet universe, scheme claims reconcile against real shops rather than disputed lists. Cross-territory flow becomes visible in a distribution management system, or DMS, the software that manages distributor stock, secondary sales and claims. Visibility here is a shared picture, never a crackdown.

How do reps profile outlets when the signal drops?

Profiling has to work with zero signal and sync when the phone is back online, or reps will stop capturing.

This is a headline requirement, not a footnote. Data costs about 2.4% of monthly income per gigabyte in Sub-Saharan Africa, so reps ration their connection. An app that stalls in a low-coverage territory sends the team back to paper by lunchtime.

Battery matters just as much. Test any profiling form on a low-end Android, in an open market, before roll-out. Geo-verified capture also protects the rep, because a time-stamped visit is evidence when an incentive payout is questioned. Owner names and phone numbers are personal data, so capture only what reps need and follow the data protection rules of each country you sell in.

How does BeatRoute turn outlet profiles into field action?

BeatRoute is a sales force automation and distributor management platform. It captures the profile on the rep app and feeds it into visit plans, campaigns and targets.

It is a global platform tailored for African trade, which is why brands like AAVA Brands and BUA Foods run on it. AAVA Brands in Nigeria runs BeatRoute SFA, DMS and Goal-Driven AI, which sets store-level targets, and grew store-level sellouts 25 to 30%. See the AAVA Brands case study.

BeatRoute serves 200+ enterprise brands across 20+ countries, reaching 2M+ retailers and 100K+ users. Brands that run every lever on one platform see a 12.6% average first-year sales uplift (BeatRoute research).

If you cover a few hundred outlets with a handful of reps, a spreadsheet and a clear scoring rule will do. A platform earns its place when thousands of outlets, many reps and several distributors make the data hard to keep clean and current.

Where should you start?

Start with one territory, not the country. Census it, score it on real data, and rebuild the journey plan from the classes.

Segmentation is not a labelling exercise you finish. It is a loop, because every visit adds data that can promote or demote an outlet next cycle. Tolaram built Indomie's reach on 1,000 distributors and 600,000 retailers, and that network was learned outlet by outlet.

Get an instant demo and see which outlets in one territory are carrying your volume.

Frequently asked questions

What is the difference between customer profiling and segmentation?

Profiling is the data you capture on each outlet: location, format, selling space, SKU range, competitor presence, price band and share of shelf. Segmentation is the grouping you build from that data, such as Class A, B and C. Profiling supplies the inputs, segmentation decides the action.

How do you profile outlets in markets where shops have no addresses?

You run an outlet census, also called field KYC. Reps walk the territory, geo-tag each shop, photograph it, and record format, owner and category range on the app. No purchasable outlet database exists for most African markets, so the list has to be built and then kept current.

Why does rep intuition misclassify outlets?

Reps grade an outlet on how big or busy it looks. A large store that gives your category one shelf is not Class A for you. Intuition also ignores trajectory, so a smaller shop with room to double gets under-visited. Structured data makes both visible.

How many outlet classes should a brand use?

Three or four suit most brands. Each class needs its own visit plan, schemes and range. With more classes than that, reps struggle to remember the rules and the classes end up treated the same.

Can you profile outlets that do not buy from you yet?

Yes. A census or retail audit records every outlet on a route, with its type, size and the brands it stocks. Those profiles show where your coverage gaps are and which new shops deserve a first visit.

Where does order data come from when outlets buy through distributors?

From the distributors. Secondary sales, the orders a distributor ships to each outlet, give you the order history your reps cannot see. It only works if each distributor invoice carries the same outlet ID as your rep app.

How often should outlet profiles be refreshed?

Rescore classes every quarter. Refresh shelf price, SKU range and share of shelf at least that often, and sooner after a currency move. The naira fell 40.9% in 2024, which changed price bands across the trade. Outlets also open, close and relocate, so a static census decays quickly.

Does customer profiling work without an internet connection?

It has to. On BeatRoute every feature works offline and syncs when the phone reaches a network. Data costs run about 2.4% of monthly income per gigabyte in Sub-Saharan Africa, so reps ration connectivity and abandon apps that stall.

Will reps actually complete the profile forms?

Only if the form is short and the effort is rewarded. BeatRoute uses embedded gamification, where companies set key behavioural indicators to reward good data input. Keep every field tied to a decision someone will make, and drop the rest.