How to use Google search data for strategy
Defining a marketing strategy is about choosing where to invest your time and money, and why.
This has been my job for the last 10 years, from Thomas Cook to Canon: grow profitably by designing and executing the best relevant marketing mix in the context of the companies I was working with.
As a contractor or interim marketing director, you need to quickly understand how your market is structured, where the company fits, identify and prioritise the best opportunities for growth. The search market map is a process I have developed, improved and consistently used with great results with brands as different as Hertz, lastminute.com, HSBC, GoCompare, The AA, Canon and Thomas Cook.
The question I was trying to solve was to understand what were the channels and the key topics driving to a market, along with the share of voice of the main competitors in that market. At the time, Hitwise was providing some solution to this problem, but the way the search data was accessible needed some reformating.
So we created a file which provided just this: the top 50,000 keywords for a market with the share of each competitor. We then classified the different terms, and were able to size and execute an ambitious SEO project which saw Thomas Cook leapfrog its competitors and save several millions of pounds a year in paid-search.
For Hertz, my problem was actually to define the marketing mix. The same technique demonstrated both the need to invest in Brand and the gap in organic search. Given the strength of the Hertz brand and the number of agencies, this strategy paid-off almost immediately, reverting a traffic trend with the brand investment, multiplying the SEO traffic by four in two years, whilst saving 20% in marketing costs.
The map is a visual representation of the search data in the market. It can be read like a "battle ground" where we can see areas of strengths and growth opportunities.
When we overlay the size of the search opportunity with the margin or return on investment that each area represents, it can be turn quickly into a comprehensive business plan.
Search data: a much underrated source of insights
Why does this technique work on all companies and markets? The reason is 63,000: the number of search that Google receives every second, or more than 3.5 billion in a day.
We query Google for information, but also as we shop or look for services, and, as a matter of fact, because search has become such a key element in our everyday life, it has also become the main channel driving traffic to most websites.
The following chart, based on SimilarWeb data in May 2020, is compelling: for about 50 websites in both France and the UK, 90% of their traffic was coming from just two channels: Direct and Search.
Search alone represents between 50% and 70% of the total traffic to most websites, and is the arch-dominant channel for acquisition.
With 50% to 70% of the internet traffic to most websites, the search data is the most powerful market research tool on the planet, yet it is used almost exclusively for tactical optimisation by specialist search teams.
The market map uses the same data as the search terms in companies and agencies. The main difference however is in the approach: companies and agencies monitor a generally small numbers of terms (100 to 2,000) for which the company already ranks. For paid-search, the suggestion of new keywords is provided by keyword planner or using broad match modifiers, an expensive way to do keyword research. These methods explore the vicinity of the search universe that the business is already visible for, providing only marginal growth opportunity.
The market map approach consists in collecting the search data providing traffic to the entire market, defined by the competitors operating in this market. We end up with many more terms that the company is ranking for, and, more importantly, we can benchmark the performance against other businesses.
This is an example of the ultra-competitive price comparison market in the UK, with five main players: Moneysupermarket, Comparethemarket, GoCompare, USwitch and Confused.com.
The visualisation provides immediate insights into the market: Car Insurance is the biggest category, followed by money products (mortgages, loans, credit cards, …), Broadband, Mobile contracts and Travel insurance.
We can also identify that some competitors are stronger in some categories: USwitch is dominant on Broadband, Phones and Energy for instance.
By combining the size of the category, the current share of the business, we can start identifying key opportunities.
However, we also need to estimate the difficulty that will be associated to gaining share in a specific category.
We can see that moving from car insurance to another insurance category such as home insurance or travel insurance should be "simpler" that expanding into mobile phones for instance.
This idea that some categories will be closer to others naturally leads to a spatial representation of the market, which would express the latent distances between categories and companies.
This is what the chart below is showing, providing in a glance three essential pieces of information:
the size of the business, representing how much organic search traffic they attract
the relative distances between two companies, based on how similar they are in terms of topics
consistent clusters or group of companies based on the importance of traffic they share with one another
As you can see, the spatial representation is very consistent with the real world, as map is for the physical reality:
we have four consistent clusters representing respectively: insurance and financial services (purple), mobile phones (blue), broadband (orange) and a myriad of small energy switching providers (green)
USwitch is at the center , with strong connections to all the sub-categories. However, the algorithm determined that it belonged to the broadband cluster, which is very consistent with USwitch main source of traffic as per the treemap seen previously
Among the "financial services" price comparison websites, Moneysupermarket and Comparethemarket are the closest to USwitch, which reflects the share of traffic they get from Broadband, Phones and Energy. GoCompare and Confused, with their very strong bias towards car insurance, are more distant from utilities.