Attributing purchase intent to multichannel activity is still very difficult, even if you’ve tagged up your digital universe. The internet allows us to measure so much, but you will only ever be able to record a tiny fraction of a consumer’s data. A purchase will most often be attributable to the mysterious data that includes:

  • out of home advertising.
  • conversations with family.
  • failure of a previous competing product.
  • unofficial celebrity endorsement.
  • an infinite amount of other non-trackable and personal events.

So, we need to be careful with Twitter’s claims that aiming ads on its platform at a second screen TV audience will:

  • Reduce cost per incremental acquisition from your TV advertising by 36%*.
  • Drive 8-16% more sales directionally^.
  • Give your promoted tweets 95% stronger message association. †
  • Produce 58% higher purchase intent. †

*From a marketing mix model for the UK telecoms market.
^ From a marketing mix model for 30 US Consumer Package Good (CPG) brands.
†From Nielsen surveys that poll a user’s brand associations.

Although Twitter claims this all to be true, these stats are of course, sleight of hand. That’s not to say I believe they are artificially inflated, it’s just that they aren’t statistically significant across all Twitter TV promotions (they can’t be); they are taken from modelling. And the stats produced from surveys simply ask a user what they think –measuring this can alter their behaviour, as any scientist knows. We never really know what consumers do in the wild.

And even if we accept the stats about message association and purchase intent, what, after all, is ‘association’ and ‘intent’? It sounds good, but before it leads to sales, there is another grey cloud in the path of attribution.

I love Twitter and believe in its power, it’s not wrong to advertise these stats. They are compelling and they tell a story I largely believe to be correct, even if it can’t be proven yet.

You can only decide whether or not to use promoted tweets alongside TV ads (or without TV ads, targeting likely viewers of a show with Twitter’s conversion mapping) by looking at TV, your own brand, your own messaging and your own content and products.

The stats are simply to help Twitter sell.

Twitter with TV: ok then, what is best practice?

Tweets are directly attributable to TV shows. 40% of UK peak time Twitter traffic is about TV.

These are fairly indisputable facts. People tweet before, during and after transmissions.

Twitter launched both TV ad and conversation targeting this year, for advertisers to broaden the reach and power of their ads, or to target ongoing conversations.

Brands can now reach a socially engaged audience to extend their TV messages, build upon sponsorships or use programmes as a proxy for interests and demographics.

That’s a clear opportunity. All makes sense so far.

Over the last few months advertisers such as Betfair, Dominos, and British Sky Broadcasting have used conversation targeting to reach millions of Twitter users.


BSkyB has struck a deal with Twitter to show clips from European Champions League matches as promoted tweets in real-time, during the games. This will help to promote NOW TV, Sky’s new contract-free online service.

This will be incredibly powerful because it hits three nails on the head:

Content: Real-time replays are incredibly highly valued by the sports fan and football is still tightly controlled by broadcasters, meaning it’s hard for the football fan to find these clips when they want them. So these messages are more than advertising, they provide a service to the sports fan.

Audience: Twitter will provide Sky with the right audience for any sector, but sports are particularly relevant on Twitter, discussed more than most other subjects.

Scope of engagement: Although the Champions League isn’t year round, the games last for almost two hours, and there are many games each week. Unlike a drama, the audience isn’t always engaged, and is often using Twitter during breaks in the play.


Before showing ads on US TV, GAP released them on Twitter. The spots include Dhani Harrison and Alexa Ray Joel singing covers of their parents’ songs. The content was made available after links from Gap exceeded 500 retweets. Gap also promoted tweets tied to their own TV ad slots, as well as targeting other shows.

Celebrity and music combined: Two of the most powerful tools for engagement on Twitter. Again the content is enjoyable, not merely an ad.

Exclusivity: Driving consumers to Twitter for the exclusive is smart, as it paints the picture of Twitter as an important and breaking channel.

Twitter is asking the right questions

Twitter’s advertising site has the following copy:

How can we make Twitter more engaging for consumers as they interact with TV?

How can we help drive more TV discovery and consumption for broadcasters?

How do TV and Twitter campaigns affect consumer attitudes, awareness, purchase intent and actual sales?

Do Promoted Tweets move consumer dollars, top-line revenue, and bottom-line results?

These are all the right questions. Only by engaging with these questions as well as your own brand’s audience, content and sector specifics can you get the most out of Twitter’s powerful tie up with TV.