11 January 2018
Footfall was down in the run up to Christmas and many retailers need to find new ways to get people into their stores. So how do you do it? One way to ‘Drive to Store’ is to use mobile. But how can you make sure you are driving the right people? Noelia Amoedo, CEO, mediasmart, outlines some much needed strategies
Retailers continue to see mobile driving retail revenue and brands of all sizes are expanding their strategies in this channel as a way to boost sales and customer bases.
While desktop is still the main retail revenue driver, a recent Adobe report on retail trends suggests mobile will surpass desktop very soon. The report highlights the opportunities for retailers as being one that’s available regardless of whether they have a strong mobile or app presence.
One of the key related opportunities are drive-to-store strategies, being used as an effective way to realise the full promise of the mobile channel to generate traffic in store through location based mobile advertising. Using geo-location to drive traffic to point of sale is a perfect strategy for bricks and mortar retailers with many stores, as campaigns can be focused on point of sale with links to directions and special offers etc.
It’s also a highly effective option for those with just a few physical stores because they can send users messages based on their locations, then either drive traffic to store or route them to a mobile app or call centre retailing presence.
For advertisers with no physical presence at all, geo-fenced campaigns can still be useful, as they allow them to find their audience. Targeting can now be based on elements around a user’s location and surroundings, in addition to app context and other influencing factors. To target based on interests, you can target areas where you are most likely to find users interested in your products, for example, airports if you are looking for air travellers.
The end goal of any geo-localised campaign of this type, of course, is to send the right offer to the right user at the right time in the right place.
The technology is now in place to be able to link a specific advertising creative to a specific location. Every potential customer who is close to a particular shop will then only see the dynamic creative developed to reach potential customers around that specific shop. The right offer can be delivered using relevant images, animations, videos or native ads. Then, granular reporting can be made available immediately, showing impressions, clicks and post-click engagement at POS level.
You can locate the right users based on how they are connected to the network (cellular versus Wi-Fi), to identify those that are on the move, and on other targeting parameters, such as what audience they are part of, or the mobile operator they use, for example. And you can get to them only ant the right time, when the stores are open. with the user´s location provided by their GPS location ensuring the right placeelement is achieved.
With geo-fencing technology now allowing retail advertisers to estimate footfall, who has entered a shop and thanks to which advertising channel, results can also be measured. Retailers just need to take into consideration that all the footfall attribution tools existing today are not deterministic, but probabilistic, and they require the size of the campaign, in terms of unique users reached, to be big enough for them to work – they are always based on the comparison of a control group and an exposed group within a given location.
Beyond the immediate footfall benefits, there’s a clear opportunity for retailers to reap a range of data capture advantages through geo-fencing by using it to construct segments based on where the users move. In this way, whole user segments can be created and reached down the line, based on their location history or interactions with a retailer’s services
To summarize, geo-fencing allows retailers to:
- Target based on location, measuring and analysing impact from mobile campaigns on visits into physical stores
- Find their audience and create new audiences based on what their travels or daily commutes show about users’ interests
- Dynamically change offers based on location and time
The underlying programmatic platforms that serve geo-fenced campaigns need to base their targeting on the location information provided by inventory suppliers to the programmatic marketplace. The extensive use of applications on smartphones and programmatic marketplaces has made it possible for advertisers to scale geo-fenced campaigns to extensive levels of reach. Now, the success of such campaigns depends 100 per cent on how accurate the location is of the users they target. Using the right media buying tools is critical therefore, to ensure they only use location information when it is accurate enough.
Despite the potential returns on the drive-to-store opportunity, there are many unrealistic promises being offered around the underlying geo-fencing tech that enables it. Testing and choosing the right technology is a crucial consideration, of course, but it’s this accuracy element that lies at the heart of maximising the geo-fencing and drive to store opportunity.
Location Accuracy: The crucial element
Accurate location information can be obtained in two ways. Firstly, through IP Geolocation, offered by partners who hold a matching table between IP addresses & their potential geolocation.
Here accuracy rates are around 90%-99%, to identify country, region or city, as it is based on information registered by the company.
The second way is via Latitude / Longitude Geo-location
This is provided by the GPS on the device, and certain partners are able to access this thanks to the native integration within apps. Obtained from the mobile device it identifies the actual user. User permission is required, but accuracy is high when targeting specific areas within a city.
So, with this in mind, what can influence the accuracy of a geo-fencing campaign?
To realistically maximise the drive to store and geo-fencing opportunity and address the issues involved with accurately targeting, there are three crucial elements that internet retailers need to consider for a given geo-fenced campaign.
The first is fraud. Dishonest publishers or other middlemen in the supply chain will often send “fake” location information, that may not have come directly from the user´s GPS. You need to work with tools that can differentiate between what is fake and what is not, and certain double verification processes can counter this.
The second consideration stems from the fact that within the mobile channel, by definition, location information is only available within the GPS capabilities, and only native applications can access this information. Accuracy in the mobile web environment can therefore be very low.
Lastly, publishers implement in different ways. Some may decide to read and transmit latitude and longitude information differently. Some may cache it for a certain period of time, for example, and the length of that time will have an impact on the accuracy. For example, if a publisher stores location information for 15 minutes, the accuracy will be limited by how much a user may move in those 15 minutes. For maximum accuracy, it is therefore important to focus on advertising inventory offering the maximum level of accuracy.
We recently ran a test to measure geo-fencing accuracy. A test banner was served via a tag that tried to access the GPS in the user ́s phone. When the code tried to access the GPS, a message was shown by the user ́s device asking for permission to gather information from the GPS. An OK was given in 55% of the impressions served during the campaign. When the GPS location was provided by the user, our system compared it with the location provided in the bid request.
The test revealed two things. Firstly, that there is much higher accuracy IN-app than within mobile web.
In app vs mobile web – With In-App in 23% of cases, users were within five metres of what we received in the bid request. By comparison, in mobile web, no request was found within less than 500 metres of the user GPS location.
- Higher accuracy through tablets than smartphones – In the case of tablets 67% of the impressions showed the GPS location of the user to be within five meters of the location provided in the bid request, while in the case of Smartphones it was only 22% of the impressions.
Alongside these considerations, there are two further considerations for retail marketers here:
The first is the ability of publishers to control precision: to maximize accuracy retailers must ensure their campaign partners use trusted publishers that provide high levels of accuracy.
The second consideration is the wide variation of accuracy by supply partners. Our tests show that the distance between the location provided at media buying time and the users’ GPS position when seeing the impression ranged from 28 to 68 per cent when comparing supply partners.
So, of all the considerations surrounding the drive to store opportunity – Location is the crucial element. The tools currently available that geo-target and measure in store vary widely in terms of their precision and accuracy. The opportunity for retailers is a robust one – and addressing these considerations below will go a long way to ensuring this new era of marketing opportunity benefits e-commerce and consumers alike.