Is mobile search the future of retail? Tom Smith, head of biddable media at mporium, thinks so
The apocryphal tale was that, in London, you were never more than a few feet from a rat. In New York it’s apparently inches from a cockroach. Well with today’s smartphone penetration we’re never more than three feet away from a smartphone. Letting a device run out of battery or forgetting it in another room causes nomophobia - being without our phones is even a psychological condition now.
For brand marketers, this is a huge opportunity. Mobile use has changed the way we respond to all types of stimuli. Five years ago, we wouldn’t be watching television and then reaching for our phones to check some fact or detail. Now when our curiosity is piqued, we instinctively reach for our smartphone and fire up Google.
We tend to watch TV when we are in a relaxed state. We’re open to new ideas, we have time to explore new concepts and opportunities. There is a natural link between viewing TV and being open to either purchase or purchase research. It shouldn’t come as a surprise then that the daily ecommerce activity curve closely mirrors that of TV viewing.
But while this is a great opportunity for brands, marketers are still optimising campaigns with a desktop perspective. Advertisers look at data and trends, and then optimise using long-term historic averages. They’re generally not reacting in real-time to the stimuli that are constantly reaching consumers and driving consumer behaviour. When you are used to looking at long-term historic averages, it is hard to switch to taking a real-time perspective in accordance with fluctuations in consumer behaviour. Marketers need a new mindset.
Historical data is still useful. At mporium, we have automated the analysis of historical data to reveal anomalies, markers that tell us when and why consumers react. We look for the changes in metrics such as impressions, clicks and conversions that represent outliers. Frequently these outliers can be linked to stimuli from TV and that tells us what caused those spikes or troughs to happen.
From there, we build our hypotheses. Is there a large increase in searches for a red dress because there was a red dress on The One Show? We can create these rules at scale. We can test to see if the specific stimulus was relevant for blue and green dresses too, or was it for a specific style of dress?
Above all, it’s about showing the most relevant ad to consumers in the moment that matters, when they are most likely to convert. Inevitably, everything should be linked to campaign performance. And it’s as important to know when to bid down on keywords, as to when to bid up. If you want to deliver performance, you cannot always bid high for generic terms and you want to bid low or even exit auctions when conversion levels are low.
But what about marketers who say this is all very well for the online environment, what about offline? The reality is that the two worlds do collide and the consumer interest generated by TV stimuli can often be monetised offline. We have a really interesting case study where a real-time search campaign for a supermarket drove people in-store.
When cookery shows such as Saturday Kitchen feature an unusual ingredient, it generates a surge in consumer interest – but the majority of viewers will not purchase these ingredients online. However, when a brand we worked with increased their bid for relevant keywords in that moment, they saw significant increases in both on-line and subsequent in-store conversions. The overall effect was to increase return on ad spend (ROAS) for the campaign by over 100%.
The automotive sector provides another obvious example of TV stimuli being monetised through a combination of online and offline. Bidding up for relevant keywords during Top Gear is unlikely to result in an immediate online purchase of a car. It is more likely to result in the online booking of a test drive and the subsequent purchase of a car at a dealership, or even no direct online response and a test drive booked directly via a visit to a dealership.
The market continues to struggle with the complexities associated with attribution and it is easy to default to last click attribution. The result is an overstatement of performance for brand keywords at the expense of generics. Many brands are capped out in terms of their paid search activity for brand terms and believe that achieving desirable levels on performance for generics is either impossible or just too hard.
The reality is that searches for generic terms far outnumber those for brand terms, so if you only focus on brand terms then your volume is capped. Achieving performance on generics is possible but you need to know when to bid up and when to bid down, and that requires smart technology.
Reacting in real-time to real stimuli doesn’t end at TV. In retail, there are a lot of signals causing people to search. Influencers on social media provide strong stimuli for young shoppers. There are a lot of channels that can and should be used to trigger great, responsive, accurate, real-time search campaigns. But most importantly, brand marketers need to get comfortable with monetising generics and taking a real-time mindset.