Strategies for Improving Your Direct Mail Campaigns Using Data (Holland & Barrett)

Direct mail is a staple of the marketeers arsenal. When used well it can be devastatingly effective. When used poorly, it won’t even be opened.

Today, this landed on my doorstep.


I’m not a frequent shopped at Holland & Barrett, but I have used a few of their stores in London, and do own a Rewards Card (their loyalty card).

This is the first time they have mailed me that I can remember.

The outside of the packet was relatively inciting.

The ‘Open me to find your Reward Coupon & exclusive offers inside!‘ was a strong call to action, and the badger holding the pink sign stood out against the green background.

And so I opened up.

Upon opening, the letter expanded to a 4 page spread, pages 2-4 of which are shown below.


Pages 2 and 3 in higher resolution.


What do you notice about these offers?

Complete lack of targeting

Of the 10 ‘Exclusive in store vouchers’ shown on pages 2 and 3, 6 are clearly targeted at women:

  • Health Magazine (ok, could be for men)
  • Wellwoman Vitality Drink
  • Wellwoman Boost Tablets
  • Probio 10 Capsules (picture of woman’s stomach on front)
  • Slim-be Strawberry Flavour Food Supplement
  • Perfectil Pro Tan

3 are unisex:

  • Aloe Vera Gel
  • Glucosamine Sulphate
  • Organic Drug Free Pollen Barries

Just 1 is for men only:

  • Health for Men Magazine

In a world where data is abundant, and in a market where there is strong competition, this type of marketing is not going to cut it.

It does NOTHING to make me loyal to the brand.

This is not Extraordinary Communication. This is bland and mundane.

It certainly doesn’t want to make me go to Holland & Barrett and buy anything. There just isn’t anything I like the look of in those offers.

Turning things around

I’ll admit, I don’t know the inner workings of Holland & Barrett. I don’t know how they operate their direct marketing team, so I’m going to be making assumptions.

Also, I am only singling Holland & Barrett out because I happened to be sent this leaflet. There are THOUSANDS of companies out there doing EXACTLY the same thing, probably worse, and throwing their money away.

That said, how could we improve this piece of marketing?

1. Segment by gender

The simplest form of customer segmentation you can do is based on demographics (age, gender, location, etc.) The most basic of these is gender.

They KNOW I am a man (they called me Mr. on the address label so I’m know they do). I gave them this information when I registered for my loyalty card.

So why the hell are they advertising mainly female targeted products to me? Wellwoman!?

A marketing executive fresh out of uni, with an ounce of common sense would know that is not going to produce a great conversion rate.

Initially I thought they might have got clever, and assumed/determined through 3rd party data that I have a girlfriend, and therefore these products were aimed at her.

But I don’t think so.

Even if they had, you still wouldn’t target 60% of the offers at someone who may or may not be going out with the person opening the mail.

That said, the kinds of products they are showing me are probably what I would be interested in IF I WAS a women, suggesting they MAY have done some targeting around age (older people are more interested in vitamins than tanning and losing weight – again just a guess based on stereotypes! I don’t have the data like Holland & Barrett do). I don’t know if they have used age, or everyone gets the same direct mail.

So, moral of the story, they should have segmented by gender.

Aciton: Get rid of Wellwoman and throw in some Wellman shizzle.

Also, why advertising the Healthy Magazine AND the Health For Men Magazine. Show me one or the other based on my gender.

Action: Only advertised the Healthy for Men Magazine to me as I’m a guy.

2. Consider past purchase history

To become a brand that people care about, you have to care about the people.

This direct meal tells me that Holland & Barrett have no idea who I am and certainly don’t care about me as an individual. Therefore, if I find another place to buy health food/vitamins/etc. there is absolutely no barrier to stop me moving.

When I saw the offers, the only one that SLIGHTLY peaked my interest was the Glucosamine Sulphate, because it has been shown to have some positive benefits when you are engaging in weight training.

The other still I have no interest in.

If they had followed rule one, they would already have increased the chances of me purchasing something by replacing the 6 women orientated products with stuff for men.

But by considering my past purchase history they could have done even better.

A classic example is Tesco. These guys are the kings of vouchers in the mail.

You buy a pack of avocados and the next month you get a voucher for 50% of avocados. That is useful. That will make be go back and buy more avocados. That will increase my brand loyalty with Tesco.

So before you send out a set of vouchers, you MUST at least try and do SOME sort of customer segmentation based on what people have purchased in the past.

If you have created personas of the types of people who your brand is aimed at (if not, why not?) then you will have around 2-5 archetypes. Create a direct mail for each of these groups. No doubt each group will have different patterns of purchasing.

At a guess, I would say the guys who shop at Holland & Barrett are looking for sports supplements in order to get big or get lean (that’s two segments). The younger women are probably looking to ‘tone up’ (segment three), or become healthier by eating organic stuff (segment four). The final group is probably the older crowd who need to look after themselves as they get older and have been recommended certain supplements by their doctors/nutritionalists (segment five).

Caveat – again I am making these up. In your business you need to analyse your data to create segments, rather than guessing.

I would fall into the second segment – trying to get lean – hence why Glucosamine stood out. They would have been able to spot this by looking at the product purchase history associated with my loyalty card.

Therefore, they would have a sent me a direct mail containing protein powders, male multivitamins, BCAAs, coconut oil, etc.

Action: Create different product offers for each of your customer archetypes.

That would have got me far more engaged.

3. Advanced – Mix in other data sources such as web analytics

If you want to get really sophisticated, you can. Suppose, for example you tied my online browsing behaviour and my offline purchasing behaviour together (it is possible).

Now, not only do you know what I have bought, but you also know what I have simply LOOKED at online. Your knowledge about me is dramatically improved.

For example, if you saw that I had looked at BCAAs 5 times in the past 30 days but never purchased, then send me an offer for them and it might tip me over the edge!

The inclusion of online data also helps you determine how likely I am to buy (based on how long I stay on the site) and more accurately understand what products I am likely to buy and what type of person I am (by looking at the articles I read and the products I look at).

With that knowledge, you have the power to deliver Extraordinary Communication to me, and make be become a brand advocate.

You might even decide to only send direct mail to customers who visit buy from your shops at least twice a month, or visit the online store more than once a month. Your reach would be far smaller, but you would save money on distribution costs, and your conversion rate would be higher.


I feel as though I am picking on Holland & Barrett. I am. A bit.

But the reality is that a failure to use data effectively is common place in most businesses.

Yes, it costs more money and takes longer to implement the strategies I have outlined above. But it does return a more positive ROI, makes more of a difference to people’s lives, and as a consequence gets them coming back to buy from you more often.

Because ultimately lifetime value is what you are trying to increase. Raising the ROI or CPA of a single campaign is OK, but a campaign that increased customer lifetime value, even though it costs more initially, can be far better in the long run.

I strongly believe that the future of digital marketing revolves around using data effectively. To do this you need to understand the techniques, have the right marketing technology, a joined up approach in how your structure your business, and willingness to try new things.

Those companies that are able to do this effectively will be the mega brands of the future, and will be the ones people rave about on social media.

But for now, start simple. Get the basics right.

If you are not doing any segmentation in your email, direct mail or other communications, start with gender or age. Create some personas, the different groups of people who are likely to buy from you. Understand why they are coming to you, what challenge they are trying to solve.

And then communicate with them in a way that makes it easier for them to solve that challenge.

Ed Brocklebank (aka Metric Mogul) is an analytics and digital marketing consultant. He helps business of all sizes become more data-driven through measurement, strategy and activation. He works as a Strategic Analytics Director at Jellyfish in London, as well as delivering training on behalf of Google and formerly General Assembly.