Connect with me

View John Griffiths profile on Facebook
Follow John Griffiths on Twitter
View John Griffiths profile on LinkedIn
Best New Thinking Winner 2010

Kata No. 7

Kata No 7 What happened? a review of the tracker’s armoury

This kata is for getting you thinking about how you track marketing activity – a lot of it of course relates to the tracking of advertising buy by no means all. What I have set out to do is a kind of ranking in order of usefulness of tracking questions starting from the least useful to the vital. And to annoy you I’ve thrown together brand, product and advertising measures in the same grab bag. I should say that a lot of this comes from my reading of the Mental World of Brands – which gathers together some of the most recent research studies in particular those which have looked at how the brain processes information – I’ll avoid the trendy N word! But it rhymes with clients…..

1. Prompted brand awareness

Are at the bottom of the heap but of course we use them all the time because we have to. The trouble is they tell us so little. Just because the brand name isn’t on the tip of your tongue doesn’t mean it isn’t influencing you. Or that you’re not buying it. But once a market is a) mature and b) crowded then it is entirely possible that virtually all competing brands in the marketplace are recalled – (and perceived to be advertising) even if they aren’t. What’s the difference between 90% prompted awareness and 99% – not a lot. You can hardly set it as a target for the campaign. But we use it because spontaneous measures don’t tell us what we want to know.

2. Spontaneous brand awareness

Look its great when people spontaneously mention the brand you’re trying to track and its a good sign. It proves that the brand is probably a leader and that it holds their attention. It is also a stubborn measure to shift – unless your advertising is out of this world spontaneous brand awareness doesn’t shift very much or very often – hence resorting to prompted measures. As well as % who recalled, you can draw comparisons between brands and look at order of mention – 1st being a great result – but if the brand concerned is a category leader then this won’t change very much. And a lot of the credit comes down to how well defined the category is. My first 2 tracking studies were Guinness and milk – try framing a category to make sure these got spontaneously mentioned 1st or 2nd? What stouts are you aware of is cheating! Beer and drinks was too broad a category – so what exactly were we measuring?

3. Liking/popularity

I have written about this elsewhere. We have to take this measure seriously because some people take it very seriously indeed – in the US for example the perceived likeability of an ad is seen as key to its memorability. But I’ve got problems with likeability – its a notoriously difficult concept to pin down. Ubiquity doesn’t help us – like Antarctica just because everyone knows where it is doesn’t mean they want to go there – its not a buying indicator. And you can get the very dangerous scenario where everyone likes the ads and thinnks the product is great but faced with the product on shelf are perfectly happy to buy something else cheaper. For certain categories – eg luxury, and audiences eg teenagers the idea that a product is very popular with certain people may indeed be a huge incentive. But its not to be trusted.

4. Advertising awareness

Of course we need ad awareness measures I hear you say. How do we know the advertising is having an effect if no one remembers having seen it? I should be more specific. It IS very useful to know how many people are able to describe an ad to you in detail so it is patently obvious that they have seen it. It is even more useful if people are able to tell you what the ad was saying, in other words the connection between the creative idea and the brand idea. The trouble with ad awareness pure and simple is that you can have famous ads with lousy branding and generic ads with misattribution – being aware that a brand has advertised is only part of the story – advertising that works when the ads have been taken down are worth every penny – publicity is a very expensive way of being famous. Adstock is how well advertised brands are remembered when they are not running. But even the understanding of the communications is tricky – people will decode new advertising ideas in terms of the messages they have already received – there is a considerable lag before a new message is really understood. In summary, understanding of the creative idea and the message behind it is what is valuable. But I may think the ads are great, decode them effectively but still not be in the market.

5. Customer satisfaction

The next 3 are used by certain companies including Microsoft use as a composite to predict future success. Customer satisfaction tracking has a chequered history and it deserves to. Only last week I pronounced myself happy a service I had received when the truth was there were certain things I was quite unhappy about – what I meant was I had achieved closure on the whole thing – I considered it complete. If on the other hand they had administered a questionnaire and asked me what things I was unhappy about I would have told them and it would have come out as a very different story. Now customer satisfaction is of course tracking the delivery of a service or a product when these earlier measures have been about how effectively a brand talks about a product or service before it has been delivered – so we’re probably comparing chalk with cheese. But it is a relevant measure because if framed properly it can be a loyalty indicator. Most of the time though it is what I call grunt research because the customer is being asked to grunt uhuh it was ok – which tells us precisly nothing but keeeps CRM managers in jobs for decades. I do however recommend measuring the extent of unhappiness – I once tried to persuade the RAC to questionnaire new members after a couple of years about their bad experiences on the grounds that getting the grumbles into the open was good for the retention figures. They were horrified at unleashing all that customer dissatisfaction – my point that it was there, it wasn’t going to go away unless they did something and leaving for the AA and coming back after a couple of years was an awfully expensive way for the RAC to get a bad experience out of the customers’ system. A sorry letter might go a long way to undoing a bumpy start. Measuring extreme satisfaction is of course very valuable if you can get your customers up to that pitch of hyperbole. Aside from cults like Nike and Apple there isn’t that much euphoria about.

6. Repurchase intentions

A surprisingly overlooked meausure. Most products and services depend on more than one purchase so the intention to replace (or not) is important. The software market totally depends on willingness to upgrade often after only a matter of months. If the repurchase score starts falling then look out. It means the passion is gone but perhaps that customers are perfectly satisfied with what they have. Wanting or intending to buy again or replace is actually a very useful indicator of interest and engagement.

7. Recommendation measures

A very bold individual by the name of Frederic Reicheld has claimed in the Harvard Business Review that the claimed willingness to recommend is the primary and only measure needed for growing the business. Which makes me instantly suspicious of one size fits all answers. Word of mouth has become very trendy and in my judgement there’s a lot of sloppy thinking around the topic. But as a composite score it is an interesting one – it implies satisfaction with the product, enthusiasm for the company and intention to repurchase. And arguably the more people are talking about and recommending the product the easier it is for a company to find new customers. So plausible on the surface.

8. Response/Purchase measures

The direct marketing business of course lives or dies by these measures. And in the golden age of the 1950s this was how TV advertising briefly worked – not only did they remember the ad the next day (DAR scores) but they went to the supermarkets and cleaned out the shelves. It is of course always gratifying to have proof of the direct effect of communications. But as we know life isn’t usually like this – the advertising effect takes years to work through – even after 12 months most of the measurable advertising effect is still to come. Claimed purchase scores on a tracking study take months to shift often rising as the awareness scores are dropping off – there is of course a lag effect on claimed purchasing because respondents may not have noticed that their conscious purchasing patterns have changed. But claimed purchase is usually a good if understated indicator that the campaign has begun to pay back. But of course if you can obtain a measure of behavioural change then by all means try to do so- it is the measure with which clients have least quarrel with (as long as their tills are ringing that is!) The catch of course is that at any given time the measurable response is a minority of your audience. ehs once claimed to have carried out a mailing with 100% purchase rate but it was of course a glorious exception.

9. Personality and performance scores

These have been much vilified in recent years largely because of the doubt that has been cast over the cognitive model of how advertising works. If advertising mostly works under the radar then what on earth can be the point of asking people to rate different brands according to personality or performance scores – surely the findings are a direct product of the research method – people form their beliefs as they are being asked. What remains true is how instinctively people are able to articulate personality and performance dimensions for brands. What is less convincing is whether brands occupy upwards of 20 axes and can be precisely tracked versus their competitors. What is more likely is that every brand dimension which is a major driver to purchased is likely to be occupied by one or 2 brands who are close to owning or embodying that particular value – they function as prototypes for that value (Volvo = safe/VW = reliable). When asked to name brands which deliver against this association respondents are likely to spontaneously name only one or two brands not because they can’t remember any others but because the association weaker or absent. Which means that if you want your brand to have a differentiated positioning then you need to map the key drivers for the market. And you need to put your brand on one of them. And it needs to be first or second. Or don’t bother. Measurable strong associations with a key driver and measures which increase so that instead of being no. 2 your brand becomes no. 1 are a genuine indicator of a perception shift. But don’t try to move all 20 dimensions at once – you’ll confuse everybody. And don’t expect these measures to shift fast – often it takes a behavioural shift – ie people starting to buy increasing numbers before their personality and performance perceptions shift in response not to the communication but to the experience of using the product.

10. All things considered

The consideration question is my favourite: If you were buying today which brands would you consider? It is a kind of happy medium because it includes a lot more people than those who have bought including a large proportion of those who haven’t bought or repurchased yet but who intend to. It separates the buyers from the tyre kickers. And the reason why it is a great if not the best measure is that a consideration set includes the brands which are prototypes for the core value drivers for the purchase. It doesn’t mean by the way that all the cars on my shortlist are the reliable ones. Normally there is a spread – but the values are dominated by a small list of brands – and there are no prizes for coming second – which is why the consideration set is such a useful and brutal measure. That’s why if your brand is 5th or 6th in a category you have to consider how to move it to first or second on one or other dimension to have a chance of getting into the consideration set. There is a catch with the consideration set of course which is how reliable it is if people aren’t in the market at that precise point. But because it expresses a commitment to consider if not actually purchase, as long as respondents are real potential buyers then it does work as an intention to purchase. There are 2 exceptions to this – one where the category is frequently purchased and there is low interest/involvement – in which case promotional mechanics can become a more influential driver. The other is where there are so many drivers to purchase that customers have to create an arbitrary shortlist out of the host of differentiated brands. But consideration is a useful shorthand for what we are trying to do as purchasers – to reduce a complicated decision to more manageable proporotions. And it is a measure which is susceptible to marketing activity. People can elevate a particular driver in importance or they can elevate a particular brand in delivering that driver.

Designed by Matthew Pattman