Following the article on loyal buyers and why they should not be the focus for brand growth, I have one follow-up question: we used to be very strong among a specific target who was very loyal parents with small kids. Over the years however we have mainly focused our investment on light users with the idea to grow the brand and put much less focus on this loyal & very important volume wise target.

As a result in a period of 5 years we lost the leadership among this target and were overtaken by a competitor. Following this experience and given the fact that investment is limited, thus we need to find the best way to distribute it enabling the brand to grow, what is the rule(s) in making sure your loyal consumers stay with you – ex. % of the budget targeted at loyals etc?

The first thing that comes to mind is the trouble with attributing a decline in sales to one particular action. Beyond just your sales data, also consider other players in the market. One possibility could be a loss in leadership of that segment due to a lack of spend on your brand’s part, or perhaps increased investment from your competitors. Another thing to consider is whether the new leader for this segment has grown across the whole market, or just this group of consumers. This will help give you a better idea of whether this was a concerted effort by your competitor to take this segment, or whether you need to reconsider how you’ve executed your own strategy. If it proves to be the former, consider why this might be the case—a potential explanation is that their media mix might be reaching consumers that you are missing.

We also recommend taking a look at how you’ve executed your marketing activities over the past 5 years. Consider your media mix—is it reaching all types of buyers equally? Your focus should be on consistently getting unduplicated reach of all category buyers and minimising the time you spend off-air. In tandem with this, it’s also important that your creative and packaging is well-branded and easily identifiable as your brand in order to boost mental availability, and that your brand is being linked to relevant category entry points. Considering these points will help your marketing communications be more effective across all types of consumers, whether they be heavy or light.

The second thing we need to think about is how this loss of leadership in one segment has affected your sales overall. Losing leadership in one segment does not necessarily mean you have lost sales across the board—it could be that your brand has become stronger across more segments instead of being the leader of just one. It’s important to consider this in the context of how your brand is growing across the whole market, not just one segment.

Thirdly, the Ehrenberg-Bass message is sometimes simplified (too much) to become “light buyers are the only important ones”. However, the truth of the matter is that we advocate for all buyers—heavy buyers are great, just don’t expect them to stay heavy year-on-year.

In line with this thinking, the focus of your marketing efforts should be to target the whole market for that category, not just the light buyers (and conversely, not just the heavy buyers either). While acquiring light buyers is vital to sustained long term growth, this is not to say that heavy buyers are not worth considering—rather that light buyers are at least as important as heavy buyers. You can find more information on this here.

So, if parents with kids are a larger part of the category, they may require a greater marketing investment. It’s important to note however, that even within this segment there are still likely to be heavy and light buyers. Marketing communications directed at light category buyers are still going to be reached and be noticed by heavy category buyers (assuming you’re not excluding them). By hitting light buyers within that segment, it’s probable that you will also be helping to maintain mental availability for your heavier buyers. We also know that the largest source of new heavy buyers comes from those on the lighter side of the buying spectrum, so by reaching those lighter buyers in the segment early, you’ll likely keep them when/if they convert to heavy buying. This report goes into detail about how heavy buyers turn into light buyers and how light buyers become heavy.

Unfortunately, we can’t make any recommendations on the % of your budget that you should target at your loyals. A good starting point might be to understand the relative size of the segment, and the flow i.e. how many buyers stay heavy, year to year, and how many light buyers become heavy (as in the report linked above). After you understand how the user base of heavy/light buyers has changed over the last few years for your brand, you can benchmark this against competitors to see if there’s any considerable deviations (and if there’s anything to worry about). But, by ensuring you have well branded, consistently on-air marketing communications that reach all category buyers (heavy and light), you increase the likelihood of your marketing efforts being more effective overall.

J.M & B.P.
03 May 2021

Could the ‘0 buyers’ be rejectors of the brand?

There may be a few rejectors, but what we know for sure is that far more it is made up of people who just do not have us in their repertoire. Brand rejection is low for all brands (fewer than 1 in 10 non-buyers typically). We see this again and again. Here is a report on this: Do consumers reject brands? People do not buy our brand primarily because it is not mentally available or physically available to them. Rejection is low.

It is also important to remember that we sometimes classify people wrongly as ’non-buyers’ when in fact they are buyers, but just not in that time period. It is very easy to underestimate the magnituade and importance of very light buyers in our brand’s customer base.

This report talks about the importance of ultra-light buyers: Ultra-lights – The Unbearable Lightness of Buying.

And this report talks about the bottom 80% of customer’s contribution to our brand: The value of Pareto’s bottom 80%.

A.S.
07 August 2020

How does the Institute define heavy, medium and light buyers in its investigations?

We tend to use the NBD-Dirchlet model as a benchmark to see if a brand has the number of heavy, medium and light buyers it should have (given the particular category it plays in and its size). Any brand has a few heavy buyers and many light buyers. There are a number of sponsors reports that discuss this:

Report 73 – Ultra-lights – The Unbearable Lightness of Buying

Report 54 – Do Your Heavy Buyers Stay Heavy, and What are They Worth?

Report 42 – There is a Pareto Law – but not as you know it

Report 65 – What Makes Heavy Buyers So Heavy?

Investigating these patterns of the distribution of the “weight” of purchases led to Gerald Goodhardt’s 20:30:50 rule. This states that the 20% heaviest buyers account for 50% of purchases, the 50% lightest buyers account for 20% of purchases, and so the middle 30% of buyers account for 30% of purchases. i.e. 20:30:50 buyers accounting for 50:30:20 purchases.

R.K.
9 January 2018

How important is it to build mental availability of light buyers?

In our corporate report #73 Ultra-lights – The Unbearable Lightness of Buying we refer to research undertaken by Dr Charles Graham. The results highlight how, over longer timeframes, (such as over six-years instead of over the shorter timeframes, like 12-months or less, that are typically examined), we see that the importance of light buyers is even greater. Hence, it is very important for brands to build mental availability of light buyers; which is a much greater challenge, since they are less familiar with the brand than are heavier buyers.

We do encourage our sponsors to look at the patterns of buying behaviour we discuss using their own data. You will need continuous panel data for five to six years. As some people drop out and others come in as replacements, you will need to filter out anyone who left or was added to the panel during the timeframe set for your analysis—so that you end you end up with panel (a sub-set of the full panel) of ‘continuous responders’. You can then re-create Table 1 from the report (refer to table below) to show the category and brand level buying behaviour.

Table 1: Ultra-light buying by brand size in five years — source: Kantar World Panel

Table 1: Ultra-light buying by brand size in five years

If you are interested in further analysis being undertaken by the Institute for your specific brands/categories, please let us know and we can discuss options for future projects.

M.F.
19 December 2017

1. Is ‘light buyer’ a definition of buying of the category or a brand or both? Can someone be light buyer of category and heavy buyer of brand or vice versa?
   1. For a brand, empirically about how many consumers are expected to be light buyers / how much volume is expected to be driven by light buyers?
   2. If the pareto law is now 60/20; it suggests that 40% of the volume comes from the other 80% consumers – does it imply that every brand will have 80% light buyers?
   3. The concept of NBD is for frequency (x axis) vs volume (y volume) ; can the same hold true for frequency vs number of ppl as well? In the absence of data can I use one instead of the other?

There are light brand buyers and light category buyers. If someone is a light category buyer (e.g. only buys category once a year) they can not be a heavy brand buyer even thought they may be solely loyal to the brand. Someone can be a heavy category buyer and either a light or heavy brand buyer. Systematically we find those who buy more of a category have bigger repertoires.

All brands follow the NBD – negative binomial distribution with expected numbers of light, medium and heavy brand buyers. Given this much of the brands volume comes from these many, many, many light buyers. They are also the route to growth. We do quite a few research projects where we check to see that brands have the expected numbers that they should for a brand of their size.

We focus on the number of people as that is what matters for growth. Frequency varies systematically with pentration (law of DJ) but not much so I suspect there is not enough variation there to use that instead of people.

Here are a few links to several of our reports and articles that relate to your questions.

• Corporate Report 65: What Makes Heavy Buyers So Heavy? Do they favour you or just eat a lot?
• Corporate Report 54: Do Your Heavy Buyers Stay Heavy, and What are They Worth?
• Corporate Report 42: There is a Pareto Law – But not as you Know it.

I have a couple of questions on light vs heavy buying.

1. For a brand, empirically about how many consumers are expected to be light buyers / how much volume is expected to be driven by light buyers?

2. If the pareto law is now 60/20; it suggests that 40% of the volume comes from the other 80% consumers – does it imply that every brand will have 80% light buyers?

Yes the pareto law (as described in Sharp & Romaniuk, 2007) does imply that the lightest 80% of a brand’s customers will generate about 40% of its sales.

Pareto metrics vary somewhat by time period. Most data only cover 1-year windows so they under-report just how light (infrequent) a brand’s buyers really are; those customers who buy even less frequently than once a year.

It’s quite normal for almost three quarters of a brand’s buyers to buy it less often than once a year. And yet for these super-light buyers to generate a third of the brand’s sales.

Similarly it’s not unusual for a third of this year’s sales to come from buyers who didn’t buy the brand last year (mainly because they are so infrequent).

I was interested to see what the latest light-buyer research had found as I’m wrestling with how we build an approach for measuring business potential for innovation that recognizes the marketing laws of growth. In particular, how we translate from quantitative screening results that give us data on buying intention amongst the general population and the reality of how buyers are distributed.

I’m not exactly sure what your question is. But I can tell you that our latest research looking at long-run (6 years) of continuous panel data shows that brands’ sales are supported by some very light buyers indeed. It’s quite normal for 75% of HHs to buy less often than once per year, and for these ultra-light buyers to collectively contribute a third of purchases.

Do you have any examples of actual data of the distribution of Light and Heavy buyers for specific categories? I understand the reason why in your Report 46 for Ehrenberg-Bass Sponsors (October 2008) you say that Light buyers of a category are very important. However, we’re seeing that ~20% of our buyers (Heavy) account for 60% of our sales and “light” buyers represent very little sales. However, I would like to compare this to other categories to understand how these results compare and if we really have a different dynamic.

The Pareto report states that purchase cycle matters, and a 12 month analysis would increase polarization vs. 3 months.  However, 12 months should be ok for most CPG categories.  In my dataset using 12 months of data, heavy buyers buy brands about 10x more frequently than light buyers.  Based on the dynamics of frequency, what are your thoughts on using a shorter period for CSDs?

We generally advise using quarterly analysis for brand metrics. It makes it clearer to marketing managers that their penetration is low.

For Pareto analysis it will probably reduce the Pareto share a little bit.

In your report on Pareto share, the analysis used purchase frequency as opposed to volume, and there’s a reference that frequency did not make a big difference. Looking at my dataset, heavy buyers buy at least 2x the volume per shopping trip compared to light buyers. If I use frequency instead of volume, I get a 15pp to 20pp decrease in the share of heavy buyers. Do you believe the paper might have underestimated the Pareto Share for categories with expandable consumption like CSDs, because it used frequency instead of volume?

That the category is “expandable” has a little to do with it and that it is a stockpile-able category. Price promotions cause some fairly heavy buyers to stock up with a volume purchase, and a tiny few lights may be encouraged in to buy or to stock up as well. If the analysis is done on purchase frequency this will have little effect: the heavy buyers will still buy on their more or less regular schedule. But if it is done on volume this single purchase may be enough to put a HH into top 20% and it will substantially make the heavier buyers look much heavier than the lights (increasing Pareto share).

Ironically, while this makes heavy buyers look more important these sales make very little (or no) profit.