Case Study: How we helped a publisher use VisualDNA data to triple conversions in a targeted brand campaign

We’ve been working closely with Komsomolskaya Pravda (KP), one of Russia’s oldest and best-known newspaper publishers, to help them leverage data to increase ad yields for one of their flagship properties – Russia’s leading sports website Soviet Sport . VisualDNA data more than doubled click through rates, reducing costs per click by 40%.

Like most traditional publishers KP, who also own tabloid news site Express Gazeta, have to prove they can attract a reliable audience in order to compete with aggregators such as Yandex and Google for advertisers’ digital spend – and they recognise the importance of reliable data (both their own data, and that of their party providers such as ourselves) in capturing that spend.

VisualDNA has a rich data set in Russia.  With 120,000 quizzes completed each month, scaled over a network of 45m cookies, we are the country’s leading data provider – incredibly rich data enabling us to target niche audiences at scale. We ran two campaigns for a major sports brand on behalf of their agency Mindshare Russia; one for a running event, the other for a specific product: Football Boots.

For the event, KP created custom segments by merging the audience that feature in three separate VisualDNA segments:  “Runners”, “18-24s” and “Males” to create an audience of 18-24 male runners. For the boots, KP merged “Soccer Lovers” “Males” and “18-24s” to create an audience of 18-24 males who play football.

The results below – benchmarked against their previous, un-targeted, Run of Network campaigns – more than justify the publisher’s investment in the data.

Demand for data in advertising is growing fast: programmatic buying is growing as advertisers see a clear return on data investment: getting better value than from the random audience they’d reach from a run of network. And media agencies are getting a better understanding of how to make data perform.

Five elements in particular made this campaign successful:

  1. The right segments: runners and football fanatics in addition to demographics
  2. The right number of impressions: don’t over-target your audience
  3. The right KPIs: to help you optimise and ensure you get ROI
  4. Fantastic creative: compelling visuals that appeal
  5. The right platform: in this case, a publisher with a loyal audience

This is a great example of how publishers can use data effectively. And the opportunities are growing. Technology allows publishers not just to reach audiences on their own sites, but to reach similar audiences after they’ve left thereby increasing reach: since VisualDNA’s platform is integrated with Google’s doubleclick, itself integrated with all major ad servers, virtually any publisher can now use our data to deliver targeted campaigns for any advertiser as a managed service and meet valuable brand campaign briefs.

How publishers can combine predictive analytics with CRM to convert registered users into paying subscribers

I know I’m wasting half my marketing budget, I just don’t know which half

It’s a well-worn marketing phrase and reflects a problem many would like to solve.  The better you understand your customers, the more effective your marketing – so we’ve been looking at how to apply the profiling technology we use in advertising to CRM, to see if we could predict which ‘unknown’ customers are most likely to buy.

VisualDNA has invested heavily in personality profiling technology. Using patented visual-based personality quizzes, we are able to profile individuals online with great detail and accuracy, and subsequently use behavioural models to extrapolate this detailed knowledge onto hundreds of millions of users. We group people by various characteristics into ‘segments’ – anonymised groups that can be used by businesses in communications, usually targeted advertising, product recommendation or content personalisation.

There are other applications too, one such area being CRM. Most businesses, whether they are business-to-business, or business-to-consumer, possess a database of customers, or active leads. And those customers can be categorised in all sorts of ways. But whichever way you define your customer base, it’s notoriously hard to predict who is most likely to buy.

So we wanted to test the idea that the inference algorithm we use to build our high-performing segments could also be used to identify which customers are most likely to buy out of any given data set.

Piano Media is a company who run a subscription paywall business on behalf of a large number of publishers in multiple countries in Europe; with a massive data set of customers who fit one of three categories;

1) Unknown / un-registered,

2) Registered but not subscribed,

3) Paying subscribers.

Clearly the objective for any business operating a freemium model is, having attracted the customer in the first place, migrating your unknown and registered users into paying customers. The question is, how do you know which of your registered users are most likely to become paying subscribers? At which part of your audience do you direct your marketing?

Working with Piano, we drew a random sample of 30,000 registered non-subscribers and compared their online behaviour to that of the entire population of paying subscribers; using what we call Look-alike modelling. We did this completely anonymously, using randomly generated anonymous identifiers stored in cookies.

From here we were able to assign a similarity score to every one of the 30,000 registered users – the score representing how similar user’s behaviour is to the paying subscribers and therefore how likely we think this person is to convert to a paying subscriber.

Piano then did three mailshots, under exactly the same conditions, to three separate groups within that population

• The top 5,000 (those with highest score)

• A random sample of 5,000 users

• The bottom 5,000 (those with lowest score)

The results are hugely encouraging. On the top 5,000 Piano got a conversion rate of 1.59%, compared to 0.52% for the random group and 0.22% for the bottom group. In other words, our algorithm correctly predicted those most likely to convert, based on the theory that the online behaviour of a paid-up subscriber was sufficient to build a profile on them and use that as a means to find those with a similar profile. Moreover, we correctly predicted that users with behaviour dissimilar to paying customers are a lot less likely to convert than random users, so that marketers don’t need to spend the money and energy marketing to these.

This result has significant implications. It means we can, in theory, take any database, whether online or offline, anonymously profile anyone in that database, and use our inference and Look-alike algorithms to predict behaviour to improve the effectiveness of a company’s marketing – in this case increasing conversion more than three-fold vs a random sample and weeding out those least-likely to buy.

Find out more.