Brands’ marketing investment should be more targeted
Growth is stalling, ROI is under the microscope and NPD is on the rise. The perfect storm. Yet still, on-trade product launches often move from concept to rollout before brands know which venues, serves and activation mechanics will drive sustainable sales. A structured test-and-learn programme can establish incrementality, identify the strongest execution and create an evidence-led playbook before significant investment is committed.
The short answer
In an industry so deprived of credible data, brands must de-risk an on-trade product launch through data pilots, by defining the commercial questions upfront, testing the variables that could materially affect performance and measuring what happens at customer, venue and category level. The objective is to understand why activity worked, where it worked and whether the result can sustain at greater scale.
Key takeaways
- A pilot should answer specific commercial questions, not simply recreate a smaller version of the launch.
- Incrementality matters: product sales are less valuable if they move demand from elsewhere in the portfolio.
- Think with the customer in mind.
- Serve, venue format, staff execution and point of sale should be treated as variables to test.
- A successful controlled trial must prove its results can survive with fewer controls at scale.
- Every pilot should end with a clear decision: scale, refine or stop.
The commercial context
The on-trade is like no other FMCG market, and gives consumers access to new products, served as intended, through real world experiences. CGA by NIQ reported that 38% of consumers in Great Britain had tried a new drink while out in the previous month, while 75% agreed that the on-trade offers an atmosphere they cannot recreate at home. This creates a strong opportunity for a commercial product trial, but trial alone does not establish commercial success. A new product may sell simply because of launch-period novelty. Without the right measurement, brands often see false indicators on volume, assume more sustainable demand and scale an execution that becomes less effective when those controls are removed.
What a credible product trial needs to test and measure
Start with decisions over deliverables
Before activity is designed, the brand should define what it needs to know: whether the product is incremental, which serve produces the strongest rate of sale, which venues represent the best fit and what support is commercially justified for roll-out.
Test the variables that could change performance
A positive result alone often cannot explain which part of the plan created it. Serve strategy, price, promotion, venue type, point of sale, staff advocacy and operational complexity should be treated as variables. The output should be a playbook that shows where to scale.
Measure behaviour as well as execution
EPOS and category data show what sold, while mystery shopping, customer intercepts and operator feedback help explain why. Results need a credible baseline (such as the same period last year) and should show whether the new product created additional demand or moved sales from elsewhere in the portfolio.
Prove that the result can travel
A regional or national rollout receives less control than a small pilot. The next phase should test the playbook across a larger, more varied estate under normal trading conditions. The question changes from “Can we make this work?” to “Can operators deliver it consistently?”
Evidence and example
GAZER used this approach to help Jägermeister de-risk the on-trade introduction of Jägermeister Orange. Jägerlabs began across 10 independent free-trade venues, using three controlled trials to test incrementality, serve strategy, venue fit, point of sale and operational complexity. The resulting gold-standard playbook was then introduced across 54 managed pub group venues with fewer controls. The regional phase reproduced the core performance outcomes established through Jägerlabs, taking the programme from controlled learning to regional validation across 64 venues.
Read the Jägermeister Orange case studyOur view
Our view is that brands waste millions a year, focusing on un-tested NPD and activation concepts. Without de-risking your go to market strategy, you’re putting way too much risk on the line. We see a ‘test and learn’ approach in retail all the time, building a playbook for future roll-out. However in the on-trade, at best, this may be stage-gated through a couple of managed customers. However, this is the kind of broad brush approach that ends up costing brands money and leaving distribution stalling across the supply-chain.
What should trade marketing teams do?
- Define the commercial questions before deciding the activation.
- Agree the baseline, comparison period and data access before recruiting venues.
- Separate the variables that could materially influence rate of sale.
- Measure incrementality across the relevant brand, portfolio and category.
- Combine sales data with compliance, customer and operator evidence.
- Record operational friction, including staff time and serve complexity.
- Validate the playbook in a larger estate with fewer controls.
- Finish with an explicit scale, refine or stop recommendation.
- Share insights with their customer partners to build category and insight leadership.
The bottom line
The safest route to scale is not to remove risk completely; it is to replace assumptions with evidence. When a pilot establishes what drives performance, where the opportunity exists and what operators can deliver consistently, brands can invest in rollout with greater confidence and a stronger commercial playbook.
Related questions
What should an on-trade product trial measure?
It should measure the commercial outcomes that will inform the rollout decision. These may include rate of sale, incrementality, category impact, venue performance, serve strategy, promotional response, compliance and operational complexity. The measures should be agreed before activity begins.
How can a brand prove that a new product is incremental?
The product’s sales should be assessed alongside relevant existing products and the wider category. Comparing trial results with pre-activity and same-period-last-year data helps establish whether the launch created additional demand or redistributed existing sales.
How many venues are needed for a credible on-trade trial?
There is no universal number. The sample needs enough variation to test the commercial question across relevant venue types and trading conditions. A smaller controlled phase can establish the strongest execution before a broader phase tests whether the result is repeatable at scale.
Why test a product again after the first pilot succeeds?
Small pilots often benefit from close management, highly engaged staff and stronger compliance. A broader phase with fewer controls establishes whether the playbook remains effective under normal operating conditions and reduces the risk of scaling a result dependent on exceptional support.
What should happen after the trial?
The evidence should lead to one of three decisions: scale the proven playbook, refine specific elements and test again, or stop further investment.
Sources & methodology
This article is informed by GAZER’s design, delivery and analysis of Jägerlabs and the subsequent managed pub group trial across 64 venues. Evidence included granular venue and category sales data, pre-trial and same-period-last-year comparisons, mystery shopping, customer intercepts and operator feedback. External context is drawn from CGA by NIQ, the IPA and Kantar. Detailed commercial results remain confidential.
