Marketing agencies have long been a CMO’s best friend. In exchange for some (usually personal) perks, the CMO would fork over a percentage of their company’s budget to a firm that would essentially outsource much of the CMO’s responsibilities.
Sometimes using an agency made sense. If a company needed fractional help – with building an app or designing collateral – an agency would give the company an expensive way to scale up hourly work without needing to hire employees. That goes double if the company culture didn’t exist to hire this kinds of talent anyway. Of course using an agency means paying a significant premium over that contractor’s true hourly rate, and a responsible CEO should ask if the CMO is too lazy to source and manage independent contractors themselves. But for the sake of argument let’s pretend the CMO is doing such important things that their time is better spent elsewhere.
Many agencies are seeing growth in their digital efforts: building and supporting software for clients, and using those software products to drive marketing. These digital works come in all flavors, from mobile apps to digital media. If we look at a breakdown of agency revenues based upon source, we see digital is not only growing faster than traditional agency sources, but it’s also the largest revenue generator.
This is where things get fun.
Many of an agency’s ad-hoc software efforts will only become cheaper and easier to come by. Newer Point of Sale (POS) software, as an analogue, is being built faster, cheaper and better than legacy POS software. More computing power, more mature coding languages, and even new libraries make building apps more painless. And we cannot neglect the inevitability that deep learning artificial intelligence (AI) will be writing better software than humans in the future: simply have a non-technical person feed the AI some design parameters and voila, the software appears.
Yet this only hints at the real problem facing agencies: a lack of data and data skills.
Let’s examine how marketing is changing. In a world where merchants spending large amounts of money have no idea if they’re actually earning an ROI, marketing dollars are shifting to measurable mediums.
TV captured roughly $71.3 billion in domestic [advertising] revenues in 2016, according to eMarketer, just under digital advertising’s $72.5 billion in the IAB’s digital ad revenue report.
Here’s that eMarketer graph for you visual folks.
Digital efforts have much more traceability, including clicks, views, dwell times, and demographics to go along with them. And because digital marketing is much more surgical in its precision, those employing a digital marketer’s toolbox will be able to more specifically reach much smaller subsets of potential customers than mass media ever could.
And you know how to cook up promotions that work? You use science.
Looking at a merchant’s sales data, machines will be able to recommend, test, and measure promotions much faster than any agency can dream of. This is especially true if these machines are directly integrated to the merchant’s POS, where all the sales and customer activity is already recorded.
Now, we can kick this up a notch and eliminate a lot of the testing phases in a very simple way: by using big data. Combining data across tens if not hundreds of thousands of establishments let’s us find similar businesses with similar customer demographics. When machines sift through the data and find what works based upon some common parameters it has a much better chance of running successful marketing.
Think of the below visual as a good example. Independently, each merchant (and their respective agency) has a limited view of the world. But once you have more data describing that world, you can make much better conclusions. By combining all the pieces you can complete the puzzle.
What entity has transactional data across tens of thousands of merchants and very tight integration with the POS?
The POS itself.
Via the right partnerships and data pooling (the POS market is fragmented but finding some consolidation, at least in US hospitality), we expect the POS (or their payment processing owners) to dominate this market.
This is why agencies are at a fundamental disadvantage. In addition to lacking the culture and data talent to make use of the data, they’re lacking the underlying data. And since merchants have been historically bad at even sharing their transaction data with business partners, this puts POS at the front of the line. Of course, POS/payments companies have their own set of culture and talent challenges.
Don’t expect larger merchants to see these benefits any time soon. In an industry where being an executive mostly means having the lowest golfing handicap, these merchants will continue their relationships with their existing agency partners for all the personal perks they bring.
Instead, expect POS companies to start automating marketing and replacing agencies in the long tail with independent merchants and smaller chains. Once the proof makes itself known – and it will happen fast because machines can iterate feedback loops like no agency can – larger merchants will start coming around. This won’t happen until investors replace existing leadership, but that objectivity will swap out agencies for far superior solutions.
If it sounds like a pipe dream think again: we estimate the first phases becoming a reality for the smaller US hospitality merchants as soon 2019 or 2020.
The market demands more technical skills as data underlies our decision making. The days of buying relationships with rounds of golf will become harder if you can’t show investors how you’re delivering superior ROI. Luckily for POS, the opportunity is theirs to lose.
Add comment