How iOS Data Privacy Rules is Changing Data Mining via Traditional Engagement Channels?

Feb 10, 2022 | Marketing Effectiveness Platform
Ios

How iOS Data Privacy Rules is Changing Data Mining via Traditional Engagement Channels?

How iOS Data Privacy Rules is Changing Data Mining via Traditional Engagement Channels? 1024 683 deepak

Background to Data Privacy

The concern about data privacy is ever-growing. Consumers have realized that the belief that “Everything is free on the internet” is not true. We provide data related to our likes, dislikes political & religious inclinations, etc. All the data that we give out is used to build sophisticated algorithms that come back to us as targeted ads. While large corporations were busy building ways to collect more and more data subtly, one of the leaders in user experience decided to take the matter into their own hands. Apple's latest iOS 15 release takes the first large-scale attempted solution to give privacy control back to the users. Unless you give an app explicit permission to track you (including those made by Apple), it cannot use your data for targeted ads, share your location data with advertisers or share your advertising ID or any other identifiers with third parties.

How Biopharma views Data Privacy?

While consumers would benefit from the increased privacy, the apps and companies that rely on using this data to drive their engagement numbers are not. An example of such companies is Pharma, where Marketing teams rely heavily on user data to drive engagement and improve targeting to physicians. The marketing in pharma relies on integrated data collected in form of personal promotion, non-personal promotion, and Consumer level data. Pre COVID, a lot of marketing effort in pharma relied on personal promotion where a pharmaceutical rep went to the physician’s clinic and explained the benefits of their product. The companies also used to conduct large-scale speaker programs where a guest lecturer was invited, and other physicians attended as an audience. This personal promotion, as we would expect, had a large impact on driving sales for the brand. The non-personal promotional (NPP) channels like email, mobile alerts, etc., and consumer channels like Paid search, Facebook, etc. even though important, played a secondary role.

Post pandemic Shift in Data Dissemination

The pandemic resulted in a major shift in how pharmaceutical companies promote their drugs. The focus shifted from Personal promotion to non-personal promotion & consumer tactics. The effectiveness of these channels depends on efficient targeting. Historically the NPP data helped identify the physician’s behavior like their affinity to channel and content and the demographics data aided in understanding physician’s age, location, etc. The availability of data from different sources helps in creating meaningful segments for effective targeting. This explains why Apple’s data restriction with the new IOS is puzzling to the marketers.

Over 50% of smartphone users in the US are on iOS. The older articles from Doximity suggest that this number can be higher than 75% for the physician community. Vendor platforms like Doximity, Medscape, Veeva may not be able to report back the engagement number for a given physician if they select the option “Ask apps not to track” on the iOS. This means that the single most effective metric to measure the impact of a channel, engagement, is no longer available. Can the marketing team afford to lose such a large volume of data while measuring the effectiveness of their channels?

Many commercial analytics projects rely on the availability of promotional data being tracked to an HCP. It can affect various analytics practices covering but not limited to the following:

  1. Drug launch analytics
    The launch of a drug is a very important event for every pharmaceutical company. After years of clinical trials and millions (and potentially billions) of dollars of investment the company leaves no stones unturned to make the drug a success. It wants to study the type of HCP segments, content and channels that can work for their marketing campaigns. The iOS change will make it very difficult for them to track the engagements that can help them make these decisions. The companies earlier used to measure the engagement of similar proxy drug to plan the launch of a new drug which may again be difficult now.
  2. Affinity Monitor
    Affinity monitoring uses data from multiple clients across different marketing channels. The deliveries and engagement for a given HCP is captured and an affinity score is calculated for the channels. This helps the companies to improve targeting. As an instance, if an HCP has high affinity towards emails as opposed to mobile alerts he can be targeted more with the emails. The Affinity monitor will be affected as the capture of engagement is restricted by the update.
  3. Marketing mix and promotional effectiveness
    The measurement of channel effectiveness is widely used for budget planning. The clients run Marketing mix multiple times a year to find out the performance of various channels and make decisions based on it. In the absence of widely available engagement data, the analytics team may have to look for proxies that we discuss in the next section. The media spend and contracts of third-party vendors have historically been based on ROI and cost per engagement numbers which will be difficult to obtain. The vendor themselves may struggle to evaluate what kind of marketing works and what doesn’t.
  4. Campaign design and analytics
    Similar to assigning budget for marketing channels the brand team is interested to measure the impact of smaller campaigns. This again will be difficult with limited data.
  5. Next best action
    Next best engagement gives the sequence of channels for each HCP that can maximize sales. It measures which channels should be deployed ‘when’ that can assure maximum engagement and sales. The initial sequence of channels can be created using deliveries instead of engagement. However, NBE is dynamic and changes the next channels in line based on the most recent engagement. Hence the engagement data plays an important role when it comes to deploying the engine accurately.
  6. Omnichannel dynamic campaigns
    Omnichannel campaigns platforms combine data from various personal, non-personal and DTC channels to create marketing plans at HCP level. They also take into account things like managed care status, payer-provider parity etc. that get missed in a traditional marketing mix analysis. The campaign effectiveness relies on the accuracy and recency of the data. The engagement data plays an important role in creating such cohesive and closed bound systems. Again, the iOS can upset the analysis or make it less reliable.

Alternative Methods of Procuring Medical Data from Traditional Marketing Channels

Companies are exploring alternatives that can help them to continue being effective in targeting the right physicians. Datazymes is also collaborating with its clients to identify effective ways of handling this massive change.

  1. Using delivery instead of engagement for measuring campaigns
    The vendors have full control over the deliveries, and we can use this for planning promotional campaigns. The drawback to this approach is not hard to see. The engagement is what drives sales, at least of wasted/ill-used resources. An HCP may have decided not to engage with alerts but if we don’t have this engagement data, we can keep delivering the alerts without any significant impact in sales.
  2. Using previous year’s engagement rate number for the new deliveries
    This idea stands on the assumption that the engagement rate for a physician may not change often. Since we have the latest engagement data of the physician, we extrapolate those numbers to the current deliveries to get an estimated number of engagements. The drawback of this approach is that we ignore any changes to the engagement behavior of the physicians. We will also have to discount for change in engagement behavior due to change in content and marketing strategy by a brand which defeats the purpose of an effective campaign. It also raises a question on how long we can use the pre-iOS engagement number for the physician.
  3. Leveraging multi-year data to predict the new engagement
    Instead of using the latest engagement data, we can use weighted averages of the previous year’s engagement rate. This can help accommodate the change in engagement trends for a given physician and provide a more holistic picture of possible future engagement. The multi-year engagement data for physicians can be classified by type of campaign, content, channels used, etc. to help derive a more accurate future engagement rate. This can be used for a few months at the least to a few years at best depending on the therapy area and drug lifecycle. The limitations will increase with the passage of time. The HCP may have changed their engagement behavior, there may be evolution of channels that has not been captured in the current data, a new drug may have been launched which impact physicians’ engagement to existing channels, there may be new efficacy and safety concerns about the drugs which may undermine engagement.
  4. Club together physicians based on demographics data and engagement rate
    We can segment and club the physicians based on historical data. A percentage of physicians from these groups may opt-in for ad tracking or uses a windows system (this needs to be evaluated based on the data) and we assume that other physicians follow the same pattern.

All the alternatives are approximations and there is no substitute for real-time engagement numbers. Will the vendors be able to circumvent the iOS updates? Are there any other proxies that can be more accurate? Leave your thoughts in the comments section below.

Subscribe to Our Insights

Contact us