Shopper Insights Explained

Topics: E-Bulletins / Newsletters, Insights, Shopper

ShopAbility overview what shopper insights are, how you know if you’ve got one, what to do with it and the various shopper research methods, uses, applications that get you there. For Retail World Magazine.

One of the findings of the Shopper Marketing Industry Benchmark Survey (findings to be published in Retail World in August) was that most companies, if they had more shopper marketing resources, would devote them first to shopper insights and shopper research.

So there is a common understanding of the NEED for shopper insights, but less understanding of the scope of them or how to use them. One of the most common questions clients ask us when discussing shopper research is ‘how is it being used – do you have any case studies?’

So we’ll remedy that here with a bit of an overview of what shopper research encompasses, how it works and how to use the results.


First, let’s clarify what is merely information vs what’s an insight, because the terms seem to be used interchangeably but they’re not actually the same thing.

Information is simply that. Often quantified. Eg, ‘40% of shopping trips in Category X in Retailer Y are Stock Up shops’.

So what?

The insight is what you infer from that information (which may be added to other information for the purpose of deriving insights). Insights will often be the ‘why’ behind something. Eg, ‘so what’s really going on here is …’.

Shopper data (such as Homescan – basket penetration, AWOP, spend etc) and shopper research (attitudes and behaviours) will yield information, but it’s your job and the research agency’s job to figure out the insights – and implications (‘so what this means is … ‘).

Not all information will yield an insight (some information is just background information), although all true insights SHOULD yield an implication.

Don’t feel you have to mine every data point for an insight, you might be looking for things that aren’t there because the shopper behaviour ‘just is’.


Shopper research is attitudinal as well as behavioural, where shopper data is behavioural and/or the outputs of the behaviour. Shopper insights are most powerful when attitudes and behaviours are married to behavioural outputs. Ie, when you blend together

It’s even more powerful when you marry your shopper data and research with your consumer research (particularly usage occasions). There’s not a clear cut line from when the consumer becomes the shopper. The Shopper Marketing survey highlighted this, with 75% of respondents believing that shopper starts outside of or before the store, not just in it.

We look at shopper research as covering a mix of what we call the ‘5Ws and 5Hs’: Who, What, When, Where, Why, How, How Often, How Many, How Much, and How Long. See Figure 1.


Includes things like …

Who Key shoppers of the category areShopper segment profiling
What They buy – subcategory, product, pack format, pack size, serve size
When They buy it – times of day, days of week, seasonality
Where They buy – channel , retailer and store choice for that categoryWhere within the category layout/shelf they buyWhere do they go and not go in the store? What do they see/not see? Where are the display ‘hot spots’ ?
Why They buy it – usage occasions, missions and trip types, drivers, motivators, influencesThey don’t buy it – barriers to purchaseLikes, dislikes and preferences
How They buy – how they make decisionsPurchase decision hierarchyThey shop – browsing, degree of planning vs impulse
How Often Do they buy the category or shop the channel/retailer/store? Daily, weekly, monthly, quarterly, annually?How long is it between purchases on average? (IPI – Interpurchase Interval)
How Many People buy from the aisle, from the primary and secondary locations?Items do they buy at a time?What’s the average weight of purchase (AWOP)?
How Much Do they spend on the category/at the store/per basket  – over time? Per purchase occasion?
How Long Do they spend in store? In the aisle?

Fig 1: The Shopper Research 5Ws and 5Hs. © ShopAbility 2009.

Some of the 5Ws and 5Hs will yield tactical insights and others will provide more strategic direction. As a general rule, the Hows tend to be tactical and the Ws are strategic. This varies depending on the category and channel.

What shoppers can tell you – and what they can’t
Shoppers can tell you about what they think now – what they like, don’t like and why. They can tell you about things they prefer or would like to see (different category layouts etc).
What they can’t do is forecast the future or tell you what the category or channel drivers are. That’s your and the research agency’s job to figure out.


Good, holistic shopper research combines exploratory (qualitative) and evaluation (quantitative) methods. It combines both ‘claimed’ (what they say they do) and actual (‘what they do do’) behaviours.

This is because just running instore components won’t tell you Whys or Hows in depth (particularly in a 5 minute instore interview where you’re  in a chilled area and shoppers don’t want to hang around), and just running qualitative interviews or focus groups won’t give you How Manys, for example.

Claimed methodologies will give you things like perceptions and attitudes, why they behaved that way, frequency, trip type, who buying for, purchase and usage occasions, and intended vs actual purchase.

Actual behaviour capture methodologies will provide you with things like where they go instore (navigation), how many go where (traffic), how long they take (duration), what they do (interactions – browse vs buy), and who they are (gender, age).

Figure 2 is an outline of what types of major shopper research methodologies answer which of the 5Ws and 5Hs. (Note that it’s not exhaustive, rather it’s indicative). Quantitative methodologies like instore interviews and online surveys can be used to put numbers around what comes out of the qualitative depth interviews, particularly for Hows and Whats, so it’s not completely clear cut.


Research Type

5Ws and 5Hs covered


(includes, not limited to …)

Accompanied shops and depth interviews QualitativeClaimed All 5Ws and 5Hs except How Long – in depth Shopper types and segmentationLikes, frustrations, triggers, barriers, motivationsUsage
Focus groups QualitativeClaimed All 5Ws and 5Hs except How Long – in depth Per Accompanied Shops
Instore observations(category specific or whole-of-trip shopper shadowing) QuantitativeActual WhoWhatWhen

How Long

Gender, approx age, basket typeDwell time at fixtureTraffic to browse to buy conversions
Instore interviews(intercept or exit) QuantitativeActual and claimed Ws: who, what, why (some), where (some)Hs: how many, how often, how much, hows (some) They buy the product, category, channel, storeDegree of planningBiggest influences

Trip types and usage occasions

Online Surveys QuantitativeClaimed All 5Ws and 5Hs except How Long Per Accompanied Shops, but with numbers put around itConcept testing – layouts, pricing models
Shopper diaries(paper or online) Qual or QuantClaimed (perceptions)Actual (self recorded behaviours) All 5Ws and 5Hs Per Accompanied Shops

Fig 2: The Shopper Research Methodology Matrix © ShopAbility 2010.

Instore observations (as opposed to interviews) do not interact with the shopper. At the other end of the spectrum, techniques such as instore workshops and instore focus groups can be used to get shoppers to perform tasks and exercises – and gather perceptions and experiences from these – in real time and in more depth than an instore interview would normally yield.

Where there’s not scan data, home scan or actual data available, if you have to you can run with claimed behaviours (eg online surveys and omnibus) for things like frequency, AWOP, and spend.

Techniques like Eye Tracking, which are essentially a form of video footage taken from the shopper’s point of view so you can see exactly what they are looking at (or what they aren’t) are normally a part of a broader methodology like Accompanied Shops or Shopper Diaries. This is because just looking at the footage in isolation without commentary around what they were there to buy or what they are thinking isn’t particularly useful in itself, for how much it costs (except perhaps maybe for quantified category layouts).


Shopper behaviour and perceptions changes by retailer, and individual retailer objectives and priorities vary.

Whilst some findings will be common across similar retailers in the same channel, not all of them will be. And they certainly vary by channel, starting with who the shoppers are, to what trip type they are on (shopper mission) and degree of planning/openness to impulse, among other things.

An example of channel variation is grocery (Coles, Woolworths) vs mass merchants (Big W, Target, Kmart). In grocery around 30% of shopping trips are stock up shops. Stock up shops largely don’t exist in mass merchants, as evidenced by the small number of trolleys employed in mass (outside of Christmas). Where grocery the majority of grocery trip types fall into the Stock Up, Top Up or Dinner Tonight categories, mass merchants skew more to shopper missions like Destination, Gifting, Entertaining and Leisure/Browsing.

Within channel, behaviours also change. An example is IGA vs Coles and WW in grocery. IGA shopping trip dwell times are around only half the average trip time in Coles and WW because IGA shoppers are doing more Top Up, Dinner Tonight and Entertaining shopping trips than they are Stock Up. The trip time, missions and shoppers all skew differently.

If you do research specific to each retailer you can blend together brand supplier and retailer objectives, hypotheses and priorities. These all vary by retailer and by supplier so you need to cover all bases, yet with a tailored methodology.

The other benefit of this is that all parties are engaged upfront ,and can be confident that the results that come back will answer the key things they are interested in.

What we haven’t discussed here is specific types of shopper research based on a specific brief, eg price and promotional modelling and choice modelling, which can get pretty technical. If you’re interested in these let us know and we can discuss in subsequent articles.


So now you’ve done a piece of research answering your 5Ws and 5Hs. Now what do you do with it?

First off, do the insights derivation exercise where you look at all the data that’s come back and ask – ‘What’s really going on here?’

Then you need to determine what the implications (‘What this means is …’) arising from the insights. A good screener question to ask here is ‘Do we care?’ to make sure you’re focussing on the big hits rather than sweating the small tactical stuff.

Lastly, you need to turn the implications into an action plan (‘What are we going to do about it?’), with priorities, and allocate who is responsible for what actions. Then you and the relevant retailers/manufactures need to get together to discuss what is going to go into market, either as a trial or as an initiative that is rolled out on scale.

Sounds easy, doesn’t it?!

So, we hope that helps clarify a few points and shopper data, shopper research and shopper insights.

In our next contribution we’ll walk you through some of the key findings from the Shopper Marketing survey … interesting stuff!

Until then.