Agile Marketing for Roasters: Testing Blends the Way Startups Test Products

 

Imagine spending weeks developing a new espresso blend. You source two or three coffees, experiment with different ratios, adjust the roast profile, arrange packaging and finally introduce it to your customers.

The launch looks promising. Your team likes the coffee, a few regular customers give positive feedback, and the social media response is encouraging. But after a month, sales begin to slow. Customers who bought the first bag are not necessarily coming back, and the blend is taking up shelf space without generating the sales you expected.

This is a familiar problem in product development. Businesses often fall in love with what they create before finding out whether customers actually want it.

Coffee roasters are not immune to this. Sometimes we spend so much time trying to develop the perfect blend that we forget to ask a more important question: Is there a market for this coffee?

This is where an approach borrowed from the startup world becomes interesting.

Stop treating every new blend like a major launch

Startups rarely have the luxury of spending unlimited money on an untested product. They develop an idea, put an early version in front of customers, observe the response and make changes before investing further.

Coffee roasters can do something similar.

Instead of producing 100 kg of a new blend, printing hundreds of bags and launching a full marketing campaign, a roaster could begin with a small batch. It might be offered to regular customers, tested at a partner café or introduced as a limited release.

The purpose is not necessarily to sell a lot of coffee immediately. It is to find out whether the product deserves a bigger investment.

Think about a roaster developing a chocolate-forward espresso blend. The initial idea might be to combine Brazilian and Ethiopian Arabica. But what does the customer actually want? More chocolate and body? A little fruitiness? Lower acidity? Something that works particularly well with milk?

The roaster may have a preferred recipe, but the market might prefer something slightly different.

A small test can reveal that before the business commits to a full production run.

A blend is a business assumption, not just a recipe

Every new coffee product comes with assumptions.

A roaster might believe that customers want a premium single-origin espresso, that a particular origin will attract attention, or that people will pay more for a blend with a detailed sourcing story.

Those assumptions may be correct. But they should not be confused with established demand.

Consider two possible espresso blends:

• A 70% Brazil and 30% Ethiopia combination, designed around chocolate, nuts and a little fruitiness.
• An 80% Brazil and 20% Uganda Robusta combination, designed for body, intensity and a more affordable price point.

Both could make sense, depending on the target customer. But they address different preferences and potentially different price segments.

Rather than arguing internally about which one is better, a roaster can prepare small batches and let customers experience them.

The important thing is to ask a specific question. Do customers prefer the first blend? Does the second perform better with milk? Does the price difference influence their decision?

A useful experiment should help answer something that matters to the business.

Let customers taste before you print hundreds of bags

One of the easiest ways to test a new blend is through a small tasting session.

Invite a few regular customers, café owners, baristas or people who represent the audience you want to reach. Give them samples without immediately revealing which one is your proposed new product.

Ask what they notice. Is the coffee balanced? Is the acidity too prominent? Does it work as an espresso? Would they drink it every day, or is it something they would only try once?

You can also compare it with your existing house blend.

Blind tasting is useful because people sometimes respond to a coffee's story, origin or price before they have even tasted it. Removing those labels can help reveal what they actually think about the coffee.

But sensory preference is only part of the picture. Someone might genuinely enjoy a sample and still decide not to purchase a full bag.

That is why a limited paid release is more revealing than compliments at a tasting table.

The difference between liking a coffee and buying it again

This is probably the most important lesson for roasters.

Positive feedback does not automatically translate into commercial success.

A customer might say that a new blend is excellent, purchase one 250 g bag out of curiosity and never buy it again. Another customer might find the coffee less exciting but purchase it regularly because it suits their daily espresso routine.

Both responses tell the roaster something useful.

After a limited release, look beyond the initial sales. Did customers finish the bag? Did anyone return to buy another? Did a café partner ask for another delivery? Was the product profitable after accounting for green coffee, roasting loss, packaging and distribution?

These are much more useful questions than simply asking how many people liked the coffee.

A roaster should also be careful with repeat-purchase figures. Someone who buys a bag every six weeks should not be judged against someone who consumes coffee much faster. The buying cycle matters.

Sometimes the problem is not the coffee

Imagine that customers enjoy a new blend, but sales remain disappointing.

It would be easy to assume that the recipe needs changing. But perhaps the price is too high for the intended audience. Maybe the packaging does not communicate what makes the coffee different. Perhaps customers do not understand whether it is intended for espresso machines, moka pots or filter brewing.

Changing the coffee itself might not solve any of those problems.

This is why product testing should eventually include the complete offer: the coffee, the price, the packaging and the way it is presented.

A roaster could discover that the same blend sells better when offered in a smaller bag, when accompanied by clear brewing instructions or when positioned as an everyday espresso rather than a premium experimental release.

The lesson is simple: customers buy a product, not just a roast profile.

Small experiments can prevent expensive mistakes

There is also a financial reason to work this way.

Green coffee has to be purchased, roasted coffee loses weight during roasting, packaging costs money and unsold inventory ties up working capital. A blend that looks attractive on paper can become an expensive mistake if demand is overestimated.

A small-batch approach does not eliminate risk, but it makes the cost of learning more manageable.

It also gives the roaster an opportunity to understand how the blend behaves across different production batches. A recipe that works once is not necessarily a recipe that can be reproduced consistently throughout the year, particularly when the underlying green coffee changes.

This is especially relevant for blends containing coffees from multiple origins. Availability, crop conditions and purchase prices can change. A roaster needs to know whether the product can be maintained without compromising its intended flavour or profitability.

When should a roaster scale?

There is no universal sales figure or tasting score that tells every roaster when to launch a product.

A small neighbourhood roastery and a large wholesale operation have very different economics. Their customer bases, production capacities and acceptable levels of risk are not the same.

But before scaling, a roaster should have reasonable evidence that the coffee tastes the way it is supposed to, customers are willing to pay the asking price, production can be repeated consistently and the margins make commercial sense.

If customers like the flavour but reject the price, the positioning may need attention. If they buy once but do not return, the product may not fit their everyday needs. If the coffee sells well but becomes unprofitable as green coffee costs increase, the sourcing strategy needs reconsideration.

Sometimes the right decision is to adjust the recipe. Sometimes it is to change the packaging or price. And sometimes the most sensible decision is to stop producing the blend.

Not every coffee needs to become a permanent product.

A different way of thinking about innovation

The coffee industry often celebrates new origins, processing methods, roast profiles and creative blends. Innovation is important, but introducing something new does not automatically make it commercially valuable.

For independent roasters, the ability to experiment is one of their greatest advantages. They can work closely with customers, test small batches and learn from the people actually drinking their coffee.

But experimentation should have a purpose. Producing a different blend every month is not necessarily innovation. It may simply create more inventory and more work.

The more useful approach is to understand what customers are missing, develop something that addresses that need and test whether the market agrees.

A roaster does not need to get every blend right on the first attempt. It needs to avoid making expensive decisions based only on personal preference.

Perhaps the best new blend is not the one that receives the most compliments at launch. It is the one customers genuinely want to keep buying.