Its a frequently over-used term, but what does “optimisation” actually mean? Gavin Stirrat, global managing director at Voluum, explains why you need to rethink your understanding of one of marketings most vital concepts.

Have you ever scrolled through the newsfeed on your favourite social media site, and felt as though you are missing out? Not just on the polished, fabulous lives of your friends and family, but on the actual content that others seem to be seeing?

That’s an optimisation algorithm in action, a process that dictates not only which status updates, photos, news stories and ads should be displayed to you, but precisely how many of each of them there should be, in order to maximise your engagement and continued potential for clicking.

Optimisation – the branch of mathematics concerned with finding the inputs to a function that produce its highest output – plays a prominent role in nearly every field of science, technology and business. It has been employed in digital media and advertising almost since its inception. The principle is that you take competing variables and attempt to reach a Pareto efficient state, or the point at which no variable could be improved without weakening another: an optimal outcome.

The rise of optimisation
The term started to gain prominence in the mid 1990s, as search engines such as AltaVista, Infoseek and Yahoo grew in popularity. In 2001 a little-known startup called Google launched its search Toolbar, and SEO (search engine optimisation) took off.

For the past 15 years, optimisation in digital display has largely followed these same dark arts. Plug your variables into a black box and out come the best results, based on a predefined set of rules, which more often than not are completely unknown to the advertiser paying the bill.

Limitations
The challenge with this approach is that even the simplest functions can give rise to surprisingly complex solutions. And the optimisation algorithm is only as good as the rules it is given – which are provided by us mere mortals. For example, if you tell the algorithm to optimise towards the best performing clickthrough rate (CTR), the system could happily start spending budget on placements that are causing accidental fat-finger clicks or, worse, deliberately fraudulent sites. A badly designed algorithm would not take these additional nuances into consideration.

Things have remained this way because tech vendors have done a good job of marketing the ‘secret sauce’ that delivers the magic on their platform and buyers have believed in the power of the algorithm to achieve the right result, as it often will. However, by taking this approach alone, you only see the outcome of the optimisation, and never the triggers that caused these positive changes.

In addition, many buy-side platforms have their reporting and trafficking as separate systems. This means that if you see something interesting in reporting, you have to go back into the trafficking system to update the variables and set that variation of the campaign live. This is inefficient and its subsequently less likely that you’ll take that step. The tools have to be brought together.

At Voluum, by taking this integrated approach to our user interface, our clients are able to set up the equivalent of thousands of line items in a fraction of the time. This allows them to rethink their approach to optimisation.

Improving choice
With UK digital ad spend now at £8.6bn – 43 per cent of all advertising spend – and £2.6bn of that going on mobile, according to the Internet Advertising Bureau (IAB) and PwC, we know that display advertising is a mature market. As a result, analysts at both agencies and brands are more sophisticated and want a far more transparent view into the optimisation that is taking place across their campaigns.

We are now seeing buy-side digital advertising platforms offer this choice. The market has recently embraced the notion of programmable over programmatic trading – the idea that campaign optimisation and management is machine-assisted, rather than machine-delivered. This is an approach that Voluum endorses. It enables the buy side to learn far more about the actual changes in the metrics that have had a positive impact on a campaign.

We are also aware, however, that this hands-on approach is not for everyone. As such, we believe the industry norm will, once again, become about choice. If you wish to get under the hood of your campaign, learning and understanding newfound insights, you can. Equally, you might want a position that sits in the middle, between this tightly controlled approach and the black box of an anonymous optimisation algorithm.

A deeper insight
At Voluum, we believe this middle position is to provide a brand or performance marketer with the insights and understanding over how multiple variables impact campaign performance. In particular, areas where the campaign is performing particularly well, or areas where it is performing particularly badly. This is a quick way for the marketer to learn a lot about the campaign set-up and metrics, which then enables more sophisticated manual optimisation further along in the process.

In addition, a brand may want the tools to perform some level of buyer defined, auto-optimisation. This is where you provide the platform with a clear set of rules: e.g., I want to bid no more than $X for a conversion rate (CVR) between A% and B%, and $Y for a CVR between C% and D%. The system would go through the buying process and, when it comes across target placements hitting these metrics, react and optimise accordingly.

It is clear that in the mobile and digital advertising industry, the word optimisation is overused and often misunderstood or missold. It is unlikely it will ever go away. However, we believe the idea of optimisation will evolve and, in turn, advertisers will gain a greater understanding of how and why their campaigns are performing the way they are. At Voluum, with our completely open and fully transparent platform, we remain committed to driving this change and choice for advertisers forward.

This sponsored article first appeared in the June 2016 print edition of Mobile Marketing. You can read the whole issue here.