Digital Advertising: How to Fix It

In the final part of our follow up to the ‘Digital Advertising is Broken: How to fix it’ Webinar of December 16th, we look at the 5 ways in which brands can better enhance their campaigns, and ensure that they don’t make the same mistakes as the rest. We’ve previously looked at the strengths of the ad industry, and also the problems with ad techniques today. Here we share our solutions...

Audience identification

Brands are often guilty of using audience targeting incorrectly.

Sometimes they’re guilty of being too broad – we were once asked to target UK parents (of which there are 13.7m), when in reality the client wanted to target parents of teenagers living in a number of specific UK postcodes. 

Some clients hide behind data and don’t try and build a picture of their audiences. Using advanced analytics such as Quantcast where there are hundreds of audience behavioural and demographic variables is great, but if that’s all you use, it can be too narrow an aperture.

Others still rely solely on demographic or panel data with tools such as TGI, which, when used in isolation, can be misleading, as individuals in the same age, gender, sexual orientation, education level, postcode & household income categories can still have wildly different interests and purchasing habits.

We always recommend that brands understand their audiences holistically, which is why we carry out 360° audience identification utilising a comprehensive combination of panel, analytics, Quant & Qual research which incorporates their behaviours, their attitudes and their demographic information


When we speak to new clients, we’re always interested to find that many of them haven’t taken the time to truly analyse their brand, and what they stand for. Sometimes they’re guilty of trotting out their corporate vision of strapline, mantra-like, without stopping to evaluate what they really stand for and what they want to represent. If they can’t distil their brand essence and communicate it in our chemistry meeting, there’s a fairly high chance that their target audiences are misinterpreting what it means as well. When we undertake workshop sessions with clients for them to understand or create a new brand identity, the results are always revealing, and they underpin all of our creative and strategic thinking thereafter. The brand vision permeates through every layer of their products and services, and makes them evaluate what it is they’re selling. Are they selling a service, or a peace of mind solution? Are they selling a product or a tool to make their consumers’ lives easier?


Emotional Creativity

When David Ogilvy published his ‘Confessions of an Advertising Man’ in 1963, he included his 38 rules in creating advertising that sells. I often wonder how many of these rules still apply. There is an Emotional creative school of thought that has delivered some of the best ad concepts of recent years. The concept of making someone smile, laugh or cry will endear them to your brand, irrespective if they’re in ‘buying mood’ or not. This approach has undeniably proven very effective for brands such as John Lewis recently. The annual Christmas ‘Sadvert’ competition has almost been John Lewis’s undisputed domain for years, and as imitation is the sincerest form of flattery, they have been pipped to the social media post by a German Supermarket and a Spanish Lottery this year. Looking at Ogilvy’s rules, #2 immediately stands out as having been broken by JL:

2. Large promise. The second most important decision is this: what should you promise the customer? A promise is not a claim, or a theme, or a slogan. It is a benefit for the consumer. It pays to promise a benefit which is unique and competitive. And the product must deliver the benefit you promise. Most advertising promises nothing. It is doomed to fail in the marketplace. Promise, large promise, is the soul of an advertisement"- said Samuel Johnson

Further examination also finds some themes which are no longer correct:

32. Yes, people read long copy. Readership falls off rapidly up to fifty words, but drops very little between fifty and five hundred words.

It’s estimated that your average online media consumer has a 6 second attention span. According to the Marketing team behind Hillary Clinton’s Presidential campaign, 6 seconds is as long as it takes to swipe through a page of a social media feed. A Vine is 6 seconds.

It’s crucial that brands inspire an emotional response. People don’t like being sold to, and they also like having social currency....

Social Sharing

There is a lot of hot air spoken about ‘social’ and ‘content’ and there are some great failures of social media. Pepsi are infamous for diverting their entire marketing budget into social media with their ‘Pepsi Refresh’ project and foregoing their usual TV advertising, including their usual Superbowl slots – and taking themselves from the 2nd biggest cola brand in the world to the 3rd biggest (behind Coke and Diet Coke). And that’s not to say that their numbers weren’t impressive – they generated 80,000 ‘votes’, swelled their fans on Facebook to 3.5m, but forget to sell them all Pepsi. People don’t buy cola because they saw a 100x72 ad on Facebook, they buy it because of years of brand awareness and loyalty that’s been created, over TV, print etc, or because there’s a price promotion on the shelf when they’re in the supermarket. BUT, Social can be a fantastic way to generate reach into millions of people through sharing. Web 2.0 first, and then social media has created new consumer behaviours. Perceptions of individuals within networks are created by what they say, what they share, how they act etc. There are very different types of audience on each of Linkedin, Twitter, Facebook etc, however they all use the same form of social currency – social content. This means that people aren’t just signposting or amplifying content, but they’re also effectively endorsing and advocating it. If you can create something that’s so shareable and engaging that it emulates Android’s ‘Friends Furever’ and achieves 6.5m shares, that’s ‘earned’ media coverage that money can buy, but would cost you millions.

Social sharing is the lifeblood of a whole new wave of Publishing. Buzzfeed, Upworthy, Gawker, The Huffington Post are all titles which started as niche sites for sharing interesting news first – i.e. before the ‘mainstream media’, but which have grown to be as large and influential as many ‘Trad’ publishers. This is an example of how online advertising (on which these sites rely for funding) has flipped the whole publisher business model on its head. Those Publishers which have embraced this changing environment have prospered. Guardian & Daily Mail have prioritised their digital publications in very different ways, but the net effect is the same. Mailonline gets ¼ billion UVs per month, and Guardian almost 150m. Instead of sticking a logo on a 728x90 and hoping that it ends up on the daily mail leader from one of their blind programmatic buys, clients should create something that becomes the headline on MailOnline (other, higher quality newspapers are available) and is the destination for the resultant social shares. It may cost John Lewis £7m to shoot their TV ad, but it’s cheap when you look at the free and earned media they achieve.

D M_emotion


Probably the most obvious of these ‘solutions’ – but also one which is very easy to get wrong. Agencies are guilty of wilfully misleading clients when it comes to their performance KPIs, but that’s often because they’re backed into a corner to deliver results against irrelevant or misleading metrics. We recently had the good fortune to read a competitor agency’s report to a client which had the nugget “the performance suffered due to the competitive environment, which was due to increased competition”. Unless you have a monopoly, competition is pretty much a given, so why are clients forced to read these ‘insights’? Perhaps it’s because clients insisted on weekly reports that they’ll never read in the SLA?

There are two main problems with measurement:

1) There’s not enough trust between agencies and their clients. The reality is that when strategising, planning and buying online campaigns, there are a huge amount of unknown variables which can affect performance. In truth, most agencies don’t really know why their click through rate on MPUs went up from 0.35% to 0.4% week on week, but they usually showcase this sliver of data as evidence that they’re doing a fantastic job and that this was down to ‘optimisation’, whatever that is. I imagine that if there’s a decline in CTR, that it’s due to ‘competition’ or ‘a test and learn which didn’t work’. Clients need to allow agencies some freedom to ‘not know stuff’. They should look at results holistically over a long period of time. They shouldn’t micromanage metrics, or analyse every line on every plan. The truth is that these days most agencies add value, not in the planning or buying (the very few who are independent and not bound by trading deals are still ultimately slightly impotent when it comes to optimisation which is largely dynamic and automatic these days) but in their level of audience insights, performance analysis and insights, and strategic recommendations. So considering that most clients are instructing agencies because they a) don’t have the skills or resource in house to do it themselves and b) ostensibly hire agencies because they are specialists, then they should allow them to specialise and they should trust them.

2) Many metrics are unrepresentative of performance. A big number of millions of impressions doesn’t signify ‘reach’. A display ad CTR is not representative of ‘targeting effectiveness’. A 5 second play of a video before it’s clicked to exit doesn’t represent ‘engagement’. If you don’t have a robust attribution model it’s fairly insignificant that you’ve driven X-thousand post impression sales within your 180 day cookie window. Many clients and agencies report on their various strands of activity in silos, as if it’s they operate in vacuums. Social, display, search etc, are sometimes as integrated as separate tabs on the same spreadsheet, but they don’t highlight the overall effect of each of those ‘touchpoints’ in the convoluted customer journey. When consumers can often second or triple-screen and be engaging with TV, search, native and display ads all at the same time, it’s archaic to think that your post impression converting leaderboard can be defined as ‘good’ or ‘bad’ performance, in isolation.

Ultimately, brands need to have an honest relationship with their agencies, use material metrics – perhaps by sharing their targets – cost of revenue, cost per acquisition, overall revenue targets etc, rather than relying on misleading data to define success.


  • Know your audience. Really know them. Research them, look at multiple data forms, and challenge what you find.
  • Think about your brand and what it stands for. Really think about it. Then test it with your target audience. Then think about it again.
  • Use emotion in your creative. Make them laugh or cry, and you’ll probably get them to buy.
  • Give your target audience social currency. If they have something to share, it’ll endorse and amplify your brand.
  • Measure all of your activity holistically, in the long term. Don’t single out isolated KPIs, and trust your agency to deliver over the long term.

Good luck!




There are no comments yet - why not be the first?

Write your comment