You can’t plan a battle without a map of the territory, yet every day marketers fumble through defining their brand strategy without a map of their competitive landscape. Topographical intelligence has been one of the deciding factors in every major conflict. Making decisions without a map is like playing chess without seeing the board. Once you have a map you can see the position of the pieces and where they can move to, which is fundamentally what strategy is about.
What is a Brand?
Brands are built up of associations in consumer’s brains, that guide them towards the right purchases. Nescafe is known for cheap, instant, coffee, whereas for Nespresso its George Clooney and the pods the coffee comes in. These units of information we recognize from culture – ‘memes’ – help us sort products into categories. Marketers play to these stereotypes – consciously and unconsciously – to attract the right customers and avoid disappointing the wrong ones.
If a product departs entirely from the expectations of the product category – for example a country music bar playing hip hop – consumers will be confused and less likely to buy. If bad fit customers buy, they’ll regret their purchase, leaving a poor review and telling friends of their bad experience. There are a basket of memes that buyers in the product category expect. Some words and images work better than others, and so brands optimize towards them. However if memes fully converge on what works, then competitors all end up looking alike, and customers can no longer tell them apart.
The memes that form around a brand come from 3 places:
- Direct experience of the product or service
- Word of mouth
- Advertising & PR
There are the memes the company wishes upon its customers, and the ones that come from the reality of the product experience as well as the perceptions of others. Branding is the act of strategically building the right associations in consumer’s brains. Everything from packaging and website design to PR and advertising can act in concert to reinforce these memory structures for when it’s time to buy. Perceptions can’t deviate too far from reality, but they can be influenced with the right positioning and enough investment in the right areas.
Survival of the Fittest
When two brands occupy the same niche, the stronger competitor inevitably prevails. To survive, smaller brands must segment the market and differentiate their positioning. For example craft breweries differentiating on taste to escape extinction at the hands of the mega-breweries. They must use different memes to their competitors – colors, phrases, features, attributes, properties – that resonate better with the target audience but can’t easily be copied. It’s important to have a map of what memes are being used by competitors so you know where to match your opponents and when to do something different. “When the world zigs, zag”
Most marketers go by gut feel when defining their brand strategy. Instead you should systematically collect examples of campaigns in your industry in a swipe file, so you’re exposed to a wider selection of memes than being limited to your personal experience. Once enough samples are collected, you can systematically code the memes with tags to spot patterns, and decide more strategically what to imitate or differentiate on.
1. Building a Swipefile
Swipefiles have been popular since the 1960s, the golden age of advertising, when copywriters clipped copies of ads they liked and stored them in a file for later when they needed inspiration. With modern note taking apps this is a far easier and more accessible practice today. You’ll need to choose a source of competitor information, for example IMDB for movie posters, the Facebook Ad Library for competitor ads, or just the content of your competitors homepages.
Example: We’re trying promote our rental property in Miami, so we create a swipefile in the form of a Notion database to collect screenshots of all the popular Airbnb listings in the area.
2. Coding Memes
Inductive coding is a process for tagging text and images with consistent labels that can be compared to find patterns. Start by selecting a small sample (10%) of the assets you collected in your swipefile, and create codes (labels) that will cover the sample. Too many categories makes comparison impossible, so start broad (’quality’, ‘cost’, ‘location’) – you can always drill down later. Select a new sample and apply your codes, adding new ones or consolidating where useful.
Example: We run through our swipefile and tag each listing by if the main image shows a bed, living room or the beach, so we can decide on what image to use for our listing.
3. Identifying Patterns
The most valuable insights come from unstructured data — the patterns emerging that weren’t on your radar, are the most important to see. Once you’ve categorized your swipefile, start drawing conclusions and looking for opportunities. Combine your meme tags with performance data (if available) to see what correlates with success. What memes appeared most and least often? Are there opportunities for superior performance? Any memes you should avoid?
Example: When we include the number of reviews in our database, we have a proxy for performance. We find listings showing the living room have half as many reviews on average.
Advanced Meme Mapping
We used a simple example of an Airbnb rental property in Miami, but you can build on the complexity from there. We only mapped the main image of the listing to 3 memes, but we could add more tags based on what was in the image, including the colors, objects in the image and common design elements. You can probably already identify patterns in what words and phrases are used in the listings (’modern’, ‘X mins/blocks to the beach’, ‘parking’), and we could use those insights in creating our listing. Ultimately every part of the assets in your swipefile can be deconstructed into their component parts to learn more about what makes them tick.
The patterns we spot are a function of how much data we have, so it would make sense to expand our search to more than the first page of listings. There are also other data sources, for example other vacation rental sites, search engine results and social media posts which might surface more insights we can use for optimizing our brand. We can use web scraping and APIs to access both pubic and private data at scale, increasing the surface area that we cover, and maximizing our chances of finding unique insights.
Typically meme mapping works best when there are a lot of examples to work from, like with our Airbnb listings. If you don’t have enough examples of competitive products your meme labels will be too sparse and you won’t spot many trends. If your product category only has a handful of competitors, it can be helpful to look at trends in an adjacent category, or transfer styles from an entirely different industry that your target audience values.
It can also be fruitful to look for memes that work well in specific combinations, or memeplexes, even if they don’t perform in isolation. Ramping up the number of tags using machine learning algorithms and APIs like GPT-3 or Google Vision can help in this endeavor. If you have performance data you can also use statistical techniques like linear regression analysis to estimate the relative impact of each meme on performance.
If you can’t get performance data to join with your meme tags don’t despair – the frequency with which a meme appears is often good enough, presuming your competitors make smart decisions about what memes to use. In any event this activity should be used for hypothesis building: all insights should be validated through creative testing. Ultimately your meme map is only a guide: it can tell you the lay of the land, but you have to chart your own course.