Innovation, personalisation and experimentation: Why you might not be doing any of it right
The word ‘innovation’ must rank fairly highly on a list of the most used and abused buzzwords in digital marketing today. Every organisation thinks that they’re innovative in some form or other – but are they going about it the right way?
San Francisco-based Optimizely believes not. According to Global Head of Strategy Hazjier Pourkhalkhali – who would like to believe his company is ‘the biggest digital laboratory in the world’ – three quarters of all experimentation companies conduct revolves around simple two variation, A/B testing. This may be fine in some circumstances, but Pourkhalkhali argues organisations can go much further.
“Imagine someone was trying to figure out how their product page should help users purchase ancillary products,” Pourkhalkhali explains. “In a traditional top-down organisation, a CMO might say [they] want there to be a bar which says ‘users who bought this also bought…’, and so on. In that world, there’s only a single variation that comes out and a single direction they can test.
“These teams tend to be very unsuccessful because what we see is they have very low ownership, because the ideas come from top down, and they have a very low appetite for risk, because they only have one opportunity to be successful,” he adds. “When we contrast this with teams that have freedom… and has the opportunity to think creatively about how to solve this problem, we see two things happen. They end up taking more risk, because they have a multitude of variations they can try, [and] they have ownership over the direction of their business.
“It’s a very different way of working, and that actually pays off in a great way – we see these companies are 75% more successful in their testing than those who simply test two variations.”
Naturally, Optimizely has a wide array of tools at its disposal to help organisations conduct data-driven experimentation across a variety of media. Yet a lot of change is cultural, rather than technological – and with customers as large as Microsoft and Conde Nast among Optimizely’s roster, sometimes there is a lot of culture to change.
This ‘bottom-up’ rather than top-down approach is key, according to Pourkhalkhali. “A customer has 5000 digital marketers working there inside the organisation – so we’re working with their group CMO, all the way down to individual CMOs and the actual teams on the ground,” he explains. “What they have to figure out is how they create a culture where they are spreading the right skills across 5000 digital marketers’ jobs; how they aggregate insights across these teams; and look at these myriad of experiments being run and still figure out what insights can be gleaned and what direction they’re sailing towards.
“And we have to think at each level – how do we make sure that the CMOs are working on the right cultural transformation? How do we make sure that below the CMOs – say the marketing directors, the heads of digital – that they’re aware of the types of changes their teams are developing? How do we train the product managers and the marketing managers to learn how to apply experimentation to their job; how to evaluate experiments, how to come up with the right hypotheses?
“It’s really critical that [different departments] understand how they all work together – and they each have a custom-tailored track.”
A large part of the company’s message has previously been around personalisation, and how – like experimentation – organisations who dive in head first will most likely fail. If you’re in the habit of data evaluation, using data to make decisions on their strategic direction, and experimenting on a global scale, forget about personalisation, says Pourkhalkhali. “If they’re not at that level, then beginning to personalise becomes very difficult, and we see people really throwing darts at the wind,” he says.
“Personalisation is really powerful if people progress there in the right format. In our view, the companies who start with experimentation and evolve into personalisation are very successful at their craft. It’s really more around the cultural direction the business has to move towards to really make data part of their DNA, and to move into this genomic personalisation where eventually you can match together all the ways to evaluate your users and content.
“But you have to understand that this is an end state you progress towards, never a starting point.”
Fostering this culture of innovation and experimentation – and using benchmarks to help secure it – will be among the key points when Pourkhalkhali takes to the stage at DMWF Global on April 168. Optimizely has spent a lot of time over the past 12 months researching ‘how companies need to think around experimentation and what kind of transformation it requires from the business’, and the results will certainly make for interesting watching.
“One of the places where companies get stuck [is] they look at testing as this kind of ‘nice to have’ – figuring out the right coat of paint you should put on your website,” says Pourkhalkhali. “Companies are forgetting that innovation is not just around how they get a little more money out of their user base, but how they continue to stay relevant in a changing technological world.
“What we want to show is that experimentation is one of the best drivers to figure out what your course should be in this evolving landscape.”
- » Alibaba develops AI tool capable of writing 20,000 lines of ad copy per second
- » IPG to acquire data-driven marketing firm Acxiom for $2.3 billion
- » How to facilitate the return of storytelling in advertising
- » Delving into a brand’s social ecosystem: AI and the refinement of data
- » Why empowered marketers will own the customer experience by 2020