Measuring marketing effectiveness: ‘Bye, bye’ metrics, ‘hello’ answers
Advertising’s come a long way over the last 20 to 30 years and so too have the reporting and metrics used to measure success.
Remember the time when we all used to watch the PG Tips real monkeys sipping on their freshly-stewed brew, White Bearded Captain Birdseye tell us how tasty his fish fingers were and Czech model, Eva Herzigova, literally revealing the power of the Wonderbra (as well as causing several raised eyebrows and a few accidents in London!) – here’s that infamous shot, just to refresh your memory…
Well, that was advertising back in the 80s/90s, when iconic campaigns that are still etched in the back of our mind somewhere were made and largely measured using audience numbers. i.e. what was on TV at the time, brand tracking surveys and sales data correlation.
Contrary to popular belief, correlation isn’t always the cause of what happens…
Correlation is not always causation because sales correlation, despite its widespread use, is flawed as there are other factors that can influence sales. For instance, many an ad agency has taken a sales correlation victory only to be told later on that there was an in-store promotion or seasonal factors that would explain the spike in sales and not in fact the ad campaign – awkward!
Then, the digital age began, which didn’t just create a new breed of consumers, but meant that everything became trackable.
Digital marketing undoubtedly opened up a whole new world of evaluation to the extent that there’s pretty much an online metric for everything. From SEO reach and keyword searches, bounce rates and opt-in leads, to click-through rates, page views, sales, loyal customers and retention. Need to produce a ten-page report on PPC and engagement? No problem. Want to research every single conversation that’s taking place around your chosen topics, as well as your categories? It’s all just a click away…
The rise of digital has generated increased marketing opportunities for brands to target an individual with a bespoke message. Studies showing that the average consumer is exposed to up to 10,000 brand messages a day. And if you were to report on all of these messages, starting with just the metrics we’ve listed above, then that’s a whole lot of reporting to produce and more importantly, digest.
We refer to this information overload as ‘analysis paralysis’, where there’s just so many metrics, stats and findings to wade through that all of this detail becomes overwhelming. And you see evidence of it everywhere, in the form of those reports that may look impressive, but sit on CEO’s desks for weeks on end, who inevitably ask for the info to be diced and sliced for the key takeaways.
While it’s great to generate lots of reporting information, information overload isn’t good for anybody.
Introducing – the future of reporting
While the digital era may have created a plethora of reporting data, it’s also created pioneering techniques, utilising Natural Language Processing (NLP).
In a nutshell, NLP uses computer science, information engineering and artificial intelligence to zone in all of the interactions between data and human’s natural language.
While beforehand, the answer may be there, but hidden somewhere within one of several 90-page documents or within one of many reporting pages, NLP reporting provides answers to questions as they’re being asked, such as, ‘Why are my sales down?’, ’Why did sales peak in August?’ and ‘What are the competition doing to win market share?’
Just like the digital future that lies ahead of us, marketing evaluation is evolving at a rapid pace, with increasingly sophisticated methods. NLP redefines the reporting landscape, it disrupts what we used to think was important and shows us we don’t actually want metrics, we want answers.
Copyright © 2020