Packaging marketing audience research turns guesswork into briefs you can defend

What to take away

Start from the decision, not the method

Audience research fails when it begins with a technique. It works when it begins with a decision that has a date attached.

Write one sentence before you commission anything: we need to know X so that we can decide Y by Z. If the sentence does not name a packaging change you could actually make, the research will not earn its budget.

Typical decisions include changing a carton size, adding a second colourway, moving a claim from the website onto the pack, or dropping a premium finish that customers do not notice.

Each decision needs different evidence. A size change needs household data and parcel measurements. A claim needs wording rules and a check against published rulings. A premium finish needs a test of whether buyers notice it at all.

That is why a single omnibus survey rarely settles everything. Treat the research as several small studies, each tied to one decision.

Get the decision owner to sign the sentence. If the person who approves the packaging budget has not agreed what the study is for, the findings will be argued about rather than used.

Define the audience you can actually reach

Most packaging briefs describe the audience in adjectives: busy, eco-conscious, premium. Those words do not help a designer or a printer.

Define the audience by behaviour you can observe or buy. For a UK-facing brand that means where people shop, how often parcels arrive, and how the pack is handled before it is opened.

Household characteristics matter more than attitudes at this stage. The Office for National Statistics publishes household characteristics data covering internet access and online shopping behaviour, which is a sound starting point for sizing an England or UK audience before you spend on primary research.

Once the shape of the audience is clear, move to the questions only your customers can answer. Our guide to packaging marketing target audience in England sets out how to narrow a broad market into segments that a packaging team can act on.

Keep the definition to one page. If a printer, a designer and a fulfilment manager cannot all use it, it is still too vague.

Choose the method that fits the question

Four methods do most of the work in packaging research, and most programmes use two or three of them together.

Desk research uses published statistics, rulings and standards. It is cheap, fast and defensible, and it should always come first.

Surveys size preferences and measure how strongly people hold them. They rank options well, such as two carton designs, and explain behaviour poorly.

Interviews explain behaviour. Ten to fifteen conversations will usually surface the reasons behind a survey result, including reasons nobody thought to ask about.

Observation covers what people cannot report: how a parcel is opened, where the pack is stored, what is kept and what is binned. Even a short diary study can change a structural decision.

For the qualitative side, packaging marketing customer interviews in England explains recruitment, question design and how to avoid leading people towards the answer you want.

Use published evidence before you pay for new data

Primary research is expensive and slow. Before commissioning it, exhaust the public record.

Packaging sits inside a legal framework. The Producer Responsibility Obligations (Packaging Waste) Regulations 2007, now marked as revoked on the legislation site, set out recovery and recycling obligations for businesses that handle packaging. The packaging waste regulations remain a reference point for anyone making environmental claims about materials or recyclability.

If your pack carries a claim about performance, sustainability or value, check how similar claims have been treated before. The Advertising Standards Authority publishes published ASA rulings covering packaging and advertising claims, and reading a handful of relevant cases is cheaper than defending a complaint.

Quality and consistency claims need a different kind of evidence. BSI's guidance on enhancing customer satisfaction is useful when you are writing a specification that a supplier has to meet repeatedly, because it frames satisfaction as something you deliver consistently rather than once.

Finally, any change to packing operations brings a safety duty. The Health and Safety Executive's guidance on managing risks and risk assessment at work applies to manual handling, machinery and the layout changes that new pack formats often require.

Design a sample that survives scrutiny

Sample design is where most packaging studies lose credibility. Three choices matter.

First, who counts as a buyer. If the pack is bought by one person and opened by another, both matter, and they often need different questions.

Second, how many. A survey of 300 to 500 respondents gives usable direction for ranking options. It will not support a claim about a subgroup of thirty people, so do not report one.

Third, how recruited. Panels are fast but skew towards people who take surveys. Customer lists are more relevant but smaller, and you must handle consent and data properly.

Write the sample definition into the brief before fieldwork starts. Changing it afterwards invalidates the comparison you were trying to make.

Where the sample is not representative, adjust the way you describe results. Report the shape of your respondents honestly, so readers know whether they are looking at England-wide direction or the views of a loyal customer base.

Run surveys that rank real options

A packaging survey should present choices, not abstractions. Show two or three actual designs, at realistic size, and ask people to choose.

Avoid asking whether a design is "appealing". It invites agreement and tells you nothing about behaviour. Ask which one they would pick from a shelf, which they would keep, and which they would expect to pay more for.

Keep the questionnaire short. Ten minutes is a reasonable ceiling for a consumer panel, and completion rates fall sharply beyond it.

The tooling matters less than the structure, but the wrong platform makes analysis harder than it needs to be. Our comparison of packaging marketing survey tools in England covers question types, panel access and the reporting limits of each option.

Pilot with five people before launch. Every pilot finds at least one question that means something different to respondents than it does to you.

Turn findings into personas a designer can use

A persona is only useful if it changes a decision. If it could describe any customer of any brand, it is decoration.

Build personas from behaviour, not demographics alone. What arrives, how often, who opens it, what happens to the pack afterwards, and what would make them buy again.

Give each persona a packaging implication. One might need clear opening instructions. Another might need a pack that survives a shared hallway. A third might need a format that fits a letterbox.

Our guide to packaging marketing buyer personas in England shows how to keep personas short enough to be read and specific enough to be argued with.

Review them once a year. Household habits and delivery patterns move, and a persona that is two years old is a hypothesis, not evidence.

Keep the count small as well. Three personas that a printer can recall are worth more than eight that nobody reads.

Worked example: an England-only study on a £9,000 budget

A fictional but realistic example shows how the numbers fit together. A mid-sized England-based seller of homeware ships around 2,000 parcels a month and wants to know whether a sturdier carton with printed opening instructions justifies its extra cost.

The programme is scoped to one decision: keep the current carton or move to the sturdier format. The budget is capped at £9,000.

These are illustrative figures, not quoted market rates, and they will move with recruitment difficulty and the length of the questionnaire.

The decision rule is set in advance. If at least 60 per cent of survey respondents choose the sturdier carton and interviews show the instructions change opening behaviour, the format is adopted. If not, the current carton stays and the money goes into the label instead.

That rule is what makes the study useful. Without it, the findings become a discussion rather than a decision.

Keep the programme running

One study is a snapshot. A programme is a habit.

Collect operational data continuously: damage rates, returns reasons, complaints that mention the pack, and photographs customers send in. This costs almost nothing and tells you when a bigger study is warranted.

Repeat the survey at a fixed interval, using the same questions, so that changes are comparable. Add new questions at the end rather than rewriting the middle.

Watch the channel mix as well as the customer. If parcel volumes shift towards lockers or collection points, the pack has to survive a different journey, and last year's findings may no longer apply.

Store the raw material, not just the report. Recordings, spreadsheets and photographs let a new colleague re-check a finding, and they make the next study cheaper to run.

Our forward look at packaging marketing trends and outlook for England in 2027 covers the material, regulatory and channel shifts worth building into next year's research plan.

Common questions

How much should a small packaging research project cost?

For an England-only study tied to a single decision, a realistic range is £5,000 to £12,000, with the worked example above at £9,000. Desk research alone can be done for a few hundred pounds if you already hold complaint and returns data.

Do I need primary research if published statistics already exist?

Not always. Use published data to size the audience and set the context, then spend on primary research only for the questions the statistics cannot answer, such as which of two designs people would choose.

How many interviews are enough?

Around ten to fifteen interviews with well-chosen buyers will usually surface the main reasons behind a behaviour. Beyond that, you tend to hear variations of the same points rather than new ones.

What should the output of the research actually be?

A one-page packaging brief with the decision, the evidence behind it, the audience definition and the specification. If nobody can act on it, the study is not finished.