Online to Offline—part 2: Second step to understanding ROPO with a deterministic approachCustomer Experience 4 September 2019
In the first part of this ROPO series (catch up here in case you’ve missed it), we’ve established why the ROPO phenomenon should not be left unnoticed, and unveiled the first way to tackle this rising trend with the probabilistic approach. Yet, we’ve also seen how that method lacked accuracy, with tools that perhaps didn’t answer all your questions.
You need look no further—the deterministic approach will help you get the right answers!
How the deterministic approach differs from the probabilistic one
Rather than estimating the volume of in-store visitors based on a sample of geolocated users, the deterministic approach matches online journeys with offline data expressly given by the users in your point of sale. Since it only counts people when there is a definite connection and does not extrapolate, its reliability is unquestionable.
Instead of measuring a “branding-like” effect through the proxy of store visits, the deterministic approach enables the precise measurement of sales thanks to data collected through various means: membership card, signing up to a newsletter in-store, filling out an order or survey form…
The data will often consist of: full names, zip codes, email addresses, phone numbers, and so on. Once collected, it is generally anonymized (necessary for customer privacy) and uploaded onto an online platform. The tool then bridges the gap with the online data through a common key to identify individualized clients and allow for advanced ROPO analysis, and much more!
So how can the deterministic approach help you?
Matching data can come in handy to answer a wide range of challenges that you might encounter as a marketer:
- Analysis & monitoring
A first key benefit rising from ROPO measurement hence consists in ensuring that your online investments won’t go to waste. On-off matching and dashboarding provide insights on which campaigns are driving the most sales rather than traditional online conversions, which can be used either in ongoing monitoring phases, or in detailed post-campaign impact analysis. As a result, you may optimize your media mix to increase advertising efficiency.
ROPO tools also help you create advanced communication scenarios that were not technically achievable before. By identifying your in-store consumers online, you can create new audience groups and start personalizing your messages across your digital assets and campaigns based on user profiles. For instance, activating your offline client database enables you to retarget your customers with cross-sell or upsell offers, reach a greater range of potential customers with lookalike audiences, exclude customers from acquisition campaigns…
So what are the tools available and how do they come in handy?
While deterministic tools ultimately serve the same objective—i.e.,bridging the gap between online and offline touchpoints—they all have unique features and answer different needs, with variable complexity levels and pricing structures. Here are some examples of methods and tools available to you, by growing complexity of use.
- Main benefit: large reach for both precise measurement and activation within the walled garden.
- Tools available on the market: Facebook (Analytics, Attribution, Advanced Analytics); Google Ads Data Hub (beta)
For several years, Google and Facebook have been making their digital coverage of users available—both broad and detailed in terms of socio-demographic data—to explain the breakdown of your sales, measure the impact of your media investments, and activate audiences within their respective inventories.
Facebook has three complementary tools for these use cases:
- Facebook Analytics (equivalent to Google Analytics, with data on your Facebook pages and ads in addition to your site) is a simple and appropriate tool to understand the correlations between your on-site actions and your sales to optimize the website or select meaningful digital indicators based on their real impact on sales.
- Facebook Attribution (formerly Facebook Atlas) makes it possible to measure a large part of your digital media investments (including, to a certain extent, the ones performed outside of Facebook) and to analyze them using more advanced attribution models than last-click. This measurement tool proved its value during a project between 55 and Lacoste, having allowed us to draw strong value-added conclusions on the effectiveness of digital devices on a Christmas period operation.
- Last but not least, Facebook Advanced Analytics—similar to Google Ads Data Hub—is a searchable database that aggregates your data with Facebook or Google’s data respectively, allowing you to take these analyses a step further in terms of accuracy and statistical modeling.
However, one must not be fooled—it is in these two advertising giants’ interest to make this data available for you to properly value advertising purchases within their inventories, and limits are thus imposed. In terms of measurement, it is utterly impossible to draw granular data out of these tools, and there are some perimeter limits: Facebook tracks all channels outside Youtube, and Google does not have access to Facebook impressions… Finally, in terms of audience activation, it is once again very difficult to extract the segments from each walled garden.
- Main benefit: multichannel activation
- Also useful for: retrospective ROPO measurement and analysis
- Tools available on the market: Temelio, LiveRamp, Graphinium (FR); Datalogix, Catalina (US)
Onboarding consists in uploading the advertiser’s offline database to match it with identified cookies from a partner database. From this, you can adapt your media strategy by targeting your online ads for specific customer segments through partner DSPs, as well as personalize your website content based on customer profiles generated by the tool. In addition to matching offline clients with online users, it reports their socio-demographic profiles, allowing you to determine whether your campaigns were seen by the right audience.
Webanalytics and CRM linking
- Main benefit: media and UX monitoring and optimization
- Also useful for: multichannel activation
- Tools available on the market: native linking (e.g. Google Analytics 360 and Salesforce Sales Cloud linking), or custom CRM -webanalytics linking (i.e., Adobe Analytics data import)
This solution requires custom integrations, including in-house development for data processing and import. Yet, Google Analytics and Salesforce’s Sales Cloud can now be combined to seamlessly link advertising with sales data, and help you maximize marketing investments by identifying the actual reach and efficiency of online advertising efforts. Many benefits can be derived from this combination of webanalytics with CRM data: getting a better understanding of online campaign performances hence optimizing bidding, adapting your landing pages and setting up better ad targeting. The main advantage of this method consists in having the matched data directly in the analytics tool itself, which is convenient for custom analysis as well as for monitoring by operational teams who are used to it.
- Main benefit: advanced and automated multichannel activation, access to third-party data and reconciliation with first and second-party data
- Tools available on the market: Oracle (DMP & CDP), Salesforce (DMP & CDP), Weborama, Ysance, Adobe Audience Manager, Tealium
DMPs generate a pool of online user IDs from your digital assets and campaigns data, from which they can trace back and record consumers’ online journeys. For instance: who viewed and clicked on your online campaigns, visited your website, added your products to their cart or wish list, etc. Combined and matched with offline data (generally through a Data Broker or onboarder), this tool can address valuable use cases to personalize your message across channels in an automated way, but requires a complex set-up and significant resources.
CDPs are not so different from DMPs, the key difference being that CDPs’ DNA is rather first-party data and use cases, while DMPs are better at handling third-party data. A consequence of these specializations is that, while the online reach of DMPs tends to be larger (if used with 3P data), CDP matching will be more precise—the data will be accessible and analyzable with more granularity, and have a longer term validity for targeting and on-to-off analysis.
But it is worth noting that, of course, DMPs can handle first-party, and nothing would prevent a CDP from ingesting third-party data.
- Main benefit: advanced retrospective analysis; media and UX monitoring and optimisation
- Also useful for: multichannel activation
- Tools available on the market: custom Hadoop ecosystem based on cloud offers (Google Cloud Platform, Amazon Web Services, Microsoft Azure…)
This extensive data storage base allows for the analysis and processing of large volumes of raw data that is channelled from your company’s various data streams, whether it comes from online or offline channels: the comprehensiveness of its data record and the possibility to store PII are especially useful to conduct thorough and custom ROPO analyses.
So, while it is still impossible to know for a fact whether a customer wasn’t solely exposed, but actually influenced by online advertising before purchasing in-store, these technologies can be a blessing for marketers as they are a big step towards understanding and measuring the impact of ROPO behaviors, to adapt their marketing strategy accordingly. Yet, among all these technologies, it is key that you properly frame your context & objectives in order to choose the right one(s), so check out the third article of our ROPO series to get you started!