Tag: data partnership

Week Seven- From Mass McDonald’s to My McDonald’s

The need for change

For years McDonald’s had been communicating to their audience via broad mass media and it had worked perfectly. However, we were beginning to see a stagnating return on media investment and overall we identified the following challenges.

  • Market conditions and technology improvements have changed consumer behaviour and media consumption drastically. Running consecutive TV-spots throughout the year were no longer sufficient in order to attract and communicate with today’s customers. We had to focus on digital media.
  • McDonald’s had entered many new day-parts such as Breakfast and night open, plus they expanded into new product areas such as smoothies and frappés. But to be successful in these areas it required a different approach than the usual mass marketing.

In order to change this situation, we had to understand the consumers’ journey to purchase much better. This meant that we needed to change our entire way of thinking. We had to move away from targeting broad audiences based on age, sex and fast food consumption, and start focusing on behaviour and needs in order to become much more relevant and targeted. We needed to adapt our marketing to every single moment in our customers’ daily media usage and apply a ‘MyMcDonald’s’ approach instead of a ‘MassMcDoanld’s’ approach.

The segment study and enriching media insightsMcDonald’s Custom Audiences_The granular semantic profiles with more than 3.000 unique characteristics_Image 3

 A ‘MyMcDonald’s’ approach includes enhancing and supporting all of McDonald’s customers’ preferences and experiences, so McDonald’s became highly personal and relevant. As we only had broad customer insights an in-depth segmentation study of McDonald’s’ customers’ behaviour was conducted. This study brought new knowledge and allowed us to be even more targeted, but we wanted to take it to the next level. In order to make the identified five segments actionable and move from mass communication to one-to-one communication, we matched this information with media consumption data. We then set out to combine the segmented data with an agile digital approach in a way never seen before – transforming insights directly into digital personas (profiles) to be used in our digital programmatic planning and buying tools.

Getting personalMcDonald’s Custom Audiences_Transforming the segments into digital personas_Image 1

When serving a banner for a client who opted in, or via our data partnerships, our Data Management Platform (DMP) grabs two types of information: cookie ID of the device we served an ad on and the URL where the ad was served. Our DMP then crawls the URL, scanning it for unique keywords and phrases, identifying what makes this page unique. All findings are tied together into our world model – a semantic model displaying what the Danish population is consuming in terms of specific words and phrases from articles, blogs, commercial and non-commercial sites within the last 30 days. The findings are also tied to the specific users’ profile so each ad impression tells us something about that specific user and the Danish population as a whole.

We then take the insights from the McDonald’s segment study and use them to identify users belonging to a given segment. The Immortals, for instance, are tech oriented, so we identify users where tech terms and phrases are overrepresented compared to the world model. We also know that they are predominantly male singles. From our socio-demographic segments in the DMP we know how to identify these users as well as what makes them different from the rest of the population. The system looks at all the cookies that fulfill the constraints set forth – i.e. consuming tech, interested in gaming and other things that characterise the segment and identify what makes them unique as a segment, creating a positive sample of an Immortal.

We then compare the rest of the population to that positive sample and identify the users/cookies that correspond enough to the positive sample to be included in the segment, and through our DMP/DSP integrations we are now able to buy those segments programmatically.

 Enhancing the creative output

McDonald’s Custom Audiences_Tailoring the creative to the right segment at the right time at the right platform - No Waste_Image 4

Being able to target the desired target group in such a precise manner gives the creative agency the chance to create messages and product offerings with a better fit to that specific target group. The level of precision targeting means that the creative agency can create day parted messaging for each segment e.g. targeting Joe, the Immortal, creating a McMuffin message for morning hours, Cheeseburger for lunch time, McFlurry as the tempting afternoon snack and a Big Mac meal for dinner time. So in essence, our targeting technology enables us to fulfill the marketing dream of the right audience with the right message at the right time.

Data-driven results

McDonald’s Custom Audiences_Buying of digital activities towards the segments_Image 5 Our new approach has generated significant value to McDonald’s. Not only on a media ROI level but at a core business level as well, positively affecting sales and the ‘cost per guest count’:

  • The overall digital ROI has increased from 2.6 to 5.8 – performing better than TV!
  • Programmatic Buying has the lowest cost amongst all media groups for bringing a guest into the physical restaurants – and at an index 17 compared to traditional digital display buying – so more than five times better than traditional display advertising.
  • Through the use of data, the sales lift from Digital activities has tripled from 2013 to 2014 and digital is now generating close to 20% of the total media-driven sales.

Moreover, this approach and use of technology have become best practice globally within McDonald’s and is currently being implemented across markets.


Stay in the know

Sign up
Successfully subscribed! Thank you!