July 11, 2013

Rustling up a Brand Campaign with De Cecco

In an in-depth case study, Jeanette Taylor tells us how nugg.ad boosted brand awareness and product sales for shop-bought pasta brand De Cecco, via an online campaign

“What’s interesting about this particular case study,” says Jeanette Taylor, Commercial Director, UK & Ireland at nugg.ad, “is that De Cecco chose to use only online, so it’s a rare opportunity to see the results of an online campaign without the effects of any other media influencing the outcome.”

Addressing delegates at Figaro Digital’s Display Marketing Seminar in May, Taylor explained how in 2012 De Cecco wanted to launch a new pasta product in Denmark. Their primary objectives were to raise brand awareness and to sell the product. Choosing online as the branding medium, says Taylor, raised a number of issues.

“How do you target a food product online that is bought in a shop? How do you know who is buying pasta? How do you find them? How do you influence them to buy your brand of pasta? If you’re trying to put across a brand message, how do you know if it’s working? How do you know if you’ve achieved brand uplift? Has your brand awareness improved and how do you measure it?” Simply measuring clicks, explains Taylor, is inadequate. Knowing your user’s affinity to your product and what their purchase intentions are is what’s important in assessing brand engagement.

Out with the old

In the past, De Cecco had used traditional forms of ad campaigns including TV commercials and press advertising and their brand awareness was low. This was confirmed by a pre-test conducted by TNS Gallup before the campaign which found that brand awareness in Denmark was just 12.6 per cent. De Cecco used nugg.ad’s Branding Optimiser, which combines predictive behavioural targeting and brand engagement measurement, to meet their objectives. 

“We ran a campaign of six million display impressions over a period of five weeks using seven leading publishers in Denmark’s PBT Network,” says Taylor. “The way it works is once a media owner wishes to begin offering our solution, we gather user surfing behaviour from all users across their network and this is combined with survey data of a sample of users. This gives detailed information on things such as social demographics, product interests and affinities. From the surfing behaviour and the survey data, we create a statistical model. The statistical model is an algorithm created from hundreds of data points and using this we can then create user profiles for all users across the network. From the profiles of the users you can then target your advertising against those users who fit your required profile.

“Users come to the site, they look at pages and some of those users will fill out one of our surveys. That survey will be served at random and it’s served throughout the publisher sites across the network so it gives a good cross-section of users, and it’s served until we get a statistically significant response rate, which usually requires about 3,000 completed surveys,” says Taylor.

The survey captures a wide range of information that wouldn’t be possible to ascertain solely from user surfing behaviour, from age and gender to interests and purchase intent. The information is then used to build the statistical model, make predictions for all users and create user profiles.
“When a user comes along and doesn’t fill in a survey we can still match their surfing patterns against hundreds of data points to enable us to make a prediction,” saysTaylor. “The point is that data is put into the statistical model and it’s the model that makes a prediction on the user in real time, all the time. It’s not retargeting, it’s predicting hundreds of variables about all the users. Then it allows you to set your advertising campaign up to target against any of those variables for the users that you want to reach.”

Gauging awareness

The second part of the campaign was to enhance brand awareness. How did nugg.ad approach this?

“The measurements for brand campaigns have got to be relevant to brand objectives and so it’s not going to be measuring clicks. Clicks are not enough. Clicks are a declining commodity.” Referring to comScore’s ‘Natural Born Clickers’ research, Taylor explains that only a minor, and decreasing, percentage of the total online population are clicking on ads. “Another piece of research showed that the users who click on the ads are the users with the lowest disposable income so unfortunately they’re probably not your core audience in many instances. And in Germany, comScore found an even more alarming statistic – three per cent of users accounted for 62 per cent of all clicks. So clicks are not the answer in this case.”

What, then, should be measured?

“It’s about a user’s perception,” says Taylor. “What do they think of your brand and how can you influence that? If you like, it’s about the ‘mental click’ rather than the physical.” Brand awareness, brand affinity and purchase intent are what’s measured, quantified and optimised in brand campaigns online, explains Taylor. What’s important is whether the user’s opinion has changed. “When a user goes into a shop, are they going to buy pasta, and are they going to buy De Cecco pasta?

“At the beginning of the De Cecco campaign we asked users three short questions to measure those three things and this is what it looked like: Have you heard of De Cecco pasta? Do you like it? Would you buy it?” says Taylor. The answers they received back from 2,500 users were used to build a statistical model specifically to increase brand awareness. “The way we did that was by taking the predictive behavioural targeting and the real time measurement and matching the profile of users who are likely to be brand positive taken from users that had already responded. From their attributes and their profiles we were able to model and create a target group of users that was likely to have a similar profile and therefore be positive to the brand.”

The TNS Access Internet Panel was used before, during and after the campaign and through clickstream technology they could compare users who had seen the ads against users who had not seen the ads. “After the campaign they recorded a massive increase in brand awareness for De Cecco in comparison to the pre-test. It had gone up by a multiple of two-and-a-half times which actually represents 148 per cent increase in brand awareness,” says Taylor.

The strategy resulted in a boost to De Cecco’s brand affinity by 100 per cent and an increase of purchase intent of 53 per cent. “Crucially there were also some great offline results,” adds Taylor. “This was an online campaign but it had huge offline success in terms of sales of De Cecco pasta.  Before the campaign De Cecco were showing average growth of 23 per cent this year, compared to the previous year. During and after the campaign they were showing sales growth of 69 per cent so that is a trebling of their sales growth rate.”

As Taylor points out, online is more than just a medium for direct response performance campaigns. It can also be hugely powerful for branding campaigns. “This case study shows really powerfully that with the right tools you can have hugely successful and quantifiable results online for product launches and for branding campaigns.”

nugg.ad/en/

Article by Julia Richardson

This feature appears in Figaro Digital Issue 18: July 2013