The Impact of Privacy Changes on Digital Marketing: A Fireside Chat by Datahash

The dramatic unfolding of regulation on data privacy, the deprecation of third-party cookies, and Google’s plan to delay the release of Privacy Sandbox and the phase-out of third-party cookies to 2024 have left us with questions about the true value add of First Party Data – its role and application. While regulations are here to stay, what better way to approach the subject than with key perspectives from industry representatives on their take on the scenario that unfolds?  

Marketers have a mounting challenge to deliver on their Ad spends and show active returns, while we see Ad channels scouting for the best among first-party data solutions providers – so as not to allow their brands to lose access to data. Data analysts and first-party data vendors try their best to meet demand and the shifting nature of the situation. The bottom line is – the industry needs to recognize and address the many issues with structured solutions in Conversions API and first-party data enablement.  

Easy access to first-party data provides marketers, advertisers, and brands with the opportunity to explore the regulation-compliant, and tangible route to effective data collection. While first-party data could understood by its definition, the subject continues to post unanswered questions and topics that deserve a riveting discussion. Datahash aims to educate marketers while unearthing the benefits, perspectives, and opportunities for omnichannel marketers to make the most of a situation, while also analyzing successful use cases.   


Akash Saxena, Head of Growth (Performance & CRM) Styli, Landmark Group  

Hicham Auajjar, Chief Product & Services Officer, MENA at GroupM & Managing Director at Keyade Middle East  

Gaurav Chhaparwal, Founder and CEO, Datahash    

Session Moderator: Aditya Shrivastava, Entrepreneur in Residence, Datahash   

The third-party cookie deprecation and device id restrictions since iOS14.5 have stirred up movement among marketers and advertisers. Reports may suggest that a vast percentage of businesses are not ready for this shift and many are still dependent on third-party cookies as their way of functioning. How has the phase-out impacted your business specifically?  

Akash: The question is relevant and it has impacted our business specifically very heavily. As our primary business is set-up in Saudi Arabia, we see a lot of users on applications, and hence we have a significant business coming in from our Applications and majorly on Apps from iOS devices. Over the years, we have observed a drop in almost 30-35% in install-to-activation rate which impacts new customer acquisition – we have been unable to re-target customers. And this has a direct co-relation with ‘Customer Acquisition Cost.’ We are also unable to use hacks that separate relevant and non-relevant Ads – this is crucial to bring the user back to the platform that increases conversions. This also impacts the top line because you’re unable to get incremental new users that you are looking for and this also impacts the cost of acquisition at an overall level.  

Hicham: I think in terms of changes you have to put them into context really, so you have multiple changes. The first one is the cookie deprecation, IOS 14.5 and you also have users who are increasingly conscious about their data and how it’s been used and we also have regulations that come into play. We know GDPR and CCPA but there are regulations coming across UAE, Saudi, and Egypt from what we know so far. In terms of changes of the impact – the first bucket – the ability to target audiences now denied and have those one-to-one conversations that we were used to having. The second bucket – measurement – ability to measure and attribute conversions and last but not least – the ability to optimize data.  Because we optimized based on specific events that are measured, this impacts our ability to optimize properly.   

Gaurav: While we are not an advertiser or agency, we work very closely with channels, FB, Snapchat Google and we work closely with all of them. What we see is that if I were to Zoom out – what everybody is looking at – it has been an opportunity – last 20 years ever since Digital Marketing has come into play – it has only gone up in one direction in the share of media. As of today, Digital Marketing controls over two-thirds of the overall media spending. The primary differentiator that Digital Marketing had in its favor was these two capabilities – the ability to measure or understand what money is driving what returns. The ROAS measurement defined with Digital is in question and personalized ads the targeted that enable you to deliver a good return due to machine learning algorithms through channels – that’s taking a hit as well. With privacy changes, it’s effectively a crisis for Digital Marketing. And to quote Albert Einstien, “In the midst of every crisis, there’s an opportunity.” We realize that there’s an opportunity while working with marketers and agencies and channels. There needs to be a proper solution around this change in the privacy landscape because of privacy and that’s what we’re trying to work with advertisers and marketers and agencies very closely.  And, help them move out of this crisis.   

As marketers and agencies who are actively involved in crafting campaigns – how do you perceive the user’s request for privacy, and consent management?  

Hicham: I think it is only natural for users to aim for privacy and to want control over their data – so we don’t see anything unusual there. What we know is that people will provide personal information to companies that they trust, if they know how this data will be used and they know what they are getting in return. What it means for us and advertisers, is to know the importance of first party data strategy, is not something you can put to a second degree, it’s crucial to prioritize it.  And, when you say FPD strategy you have to take it to consent, Since you have users who are worried about Privacy and you know that you need FPD, strategically for you to activate better and to be able to measure better, it also means that you need a consent strategy. It’s ultimately for advertisers its about building trust with their customers and prospects and making sure that there is an actual value exchange with the customers. This means – I am going to ask for your information, but in return I am going to make it obvious how I am going to use your data and what you’re getting out of it. Transparency is important – users want to know the data being collected, how it is going to be used and to be re-assured on the security aspect, where it is being stored and if there is an option to opt out.  Am I going to be able to do that, if I want to delete am I going to be able to share any data with you?  It is ultimately about transparency but also security and choice. So, you as an advertiser, providing security and choice, so if tomorrow I want to change the frequency of communication as an advertiser, I should be able to go about it.  

Akash: That awareness was always there even before the IOS 14 change. If you notice even before the IOS 14 change there were almost 25 to 30% of people who were actually opting out of IOS devices, so people were already aware that they had the power to stop the amount of data that they have to share with the brands they trust. This has impacted the business and the consumer in the same way and I as a user interacted with a crucial piece of communication as I went through various digital platforms rather than getting something random at this point.  

Gaurav: I agree with Akash, transparency has to be there. But the option to opt out has always been there. The issue around privacy were brought on by the excessive use of re-targeting at a product level or the other issue is around political misuse where even news itself became personalized that it was used for swaying public opinion. When you use Whatsapp, Facebook – we have been using these services free and we understand that these have to be funded by marketing. As users we know that. The choice is whether we want to see lots of relevant ads or a few targeted relevant ones. This is an obvious choice. It was not coming from real users, probably behind it was the agenda – the transparency was always non-negotiable.   

Akash: Because of all these changes we’ve actually gone back to what we have built. Top of the funnels are coming back which actually does not drive any key KPIs that you want to drive in your business. And you will not be able to evaluate the impact of these campaigns because it is difficult to build models that give you the right understanding of what is incremental growth in organic traffic vs non-organic traffic.   

Let’s come to compliance for a bit. So far, we have GDPR and CCPA which have set down clear-cut guidelines and expectations. The regulation clearly refers to AdTech and MarTech companies as controllers and processors holding them responsible for their data activity. What’s your experience and the key challenges in adherence and adoption of these guidelines?    

Akash: Getting consent and explaining what compliance means to a user has been a challenge. A lot of people don’t pay attention to the final prints when they’re consuming a brand. They are there for the product and the services they want to get into. Apart from that, I think a lot of people don’t understand what changes an organization has to make to account for that consent. If Gaurav comes to me saying that I want to opt out of my data from my platform, I have to create systems in my organizations that has to delete and erase entire history of Gaurav and any kind of data I have collected about him. That involves a huge transformation from the business themselves. The business has to constantly improve the value add and we still have so much law but the users are not aware of what this means and how this will help them in detail.  If the awareness or consent is the key problem area – that area has not been invested in heavily.   

Gaurav: We do abide by the law and we do help marketers do that but the law itself has been excessively complex. So much so that the European Commission has decided to do a data reform bill that scraps GDPR. It changes a lot – it claims that it will save a Billion pounds for the European Commission in terms of the entire privacy subject. When GDPR came out everyone had to comply and out of fear everyone ended up implementing cookie banners. Half the cookie banners when you click do not do anything, I can tell you that because we workday in and day out with advertisers. It’s a bit more as a user, as far as the user experience is concerned – How many of us would like to have an option where we don’t see that banner. If I were to and if someone was so particular about it they could just browser in the incognito mode. Law has been created to cater to specific use cases but has been generalized to a far wider requirement of adherence which is excessively complex to implement and places much burden on marketers.  

Hicham: The more deeper in-rooted challenges – Regulation in general is positive. I think consent is on part of it, the real challenge is building a Privacy by Design ecosystem. It’s not about consent but it is about building the right governance practices within your organization first across handling data in general. And, having specifics on having the type of data that you handle is crucial to invest in governance and awareness internally and externally. It’s more about turning into privacy by design model. Do I need to collect this data and how am I going to store it? Do I have the right security measures in place to protect it and ensure it goes to the right people can access it. Its a complex topic and is a necessary one. It starts by raising awareness of privacy and thinking privacy first. 

 The key so-called negative effect of the phase-out has been the loss of identifiers like Cookie Id and Device Id? How are you overcoming these challenges?

 Akash: You have to be clear on the data you are collecting from your user and you have to be clear about where the data will be used and for what purpose. Collecting First Party Data is the primary focus for every advertiser to utilize it to help them grow and deliver metrics they want to achieve. Lot of innovation has come, earlier we used to only have FB and Google social log-ins, now people have whatsapp log-ins the phone up is your primary identifier and if you add in the phone number that becomes a powerful source for businesses to get top line activations to leverage across channels. In order to optimize your campaigns CAPI is available. Instead of relying on IOS, CAPI as a solution takes all your FPD and pushes into all this partner platforms. When there is an influx and huge amounts of data the algorithms they work on they get additional data points to optimize the campaigns in a way they are still relevant. When we implemented CAPI – we see that there was a 30-35% improvement in our ROAS – this optimization does not happen overnight. The way the systems were able to identify the users and very accurately pin point the interests – that level of accuracy was not there. It was all probabilistic models and you cannot expect the same kind of results but you can come close to what you were expecting.    

 Hicham: As I mentioned, we look at it from multiple angles, we rely heavily on targeting. The second point is more on MarTech assessment – how reliant are you on Third Party Cookies as an organization. The third bit is around data assets or governance – are you data rich, what data do you own and how do you use it? And then the last bit, the measurement assessment. One is to work on your FPD strategy and what is your identity infrastructure – this is an important aspect. Data diversity becomes important – use as many signals as you can. It’s crucial and the increasing importance of data science. Modeling is another aspect that is becoming increasingly important. We also have other things that are important in how we collect data and then if we look at measurement separately a lot of things will change. MMMS will become more important, probabilistic measurement will be more important. If we look at where we are today CAPIs are definitely important to work on server-side communication. 

 Gaurav: Web CAPI or App CAPI there is a possibility to unlock the potential for omni-channel businesses specifically bringing in offline CAPI. There is a notion of lead conversions using first party data and there is Custom Audience. There are FPD techniques and all of them have most of the offerings which are compliant but there is a continued research angle to it. MMM plus MTA combinations. There are notions of Privacy Sandbox that Google is working on, multi-party communications, on device learnings which does not share the data but it happens on the device itself. 

 Tell us about your experience with the consented First Party Data-based tracking?

 Hicham: What I can tell you specifically is we look at it in two ways – Measurement what is the outcome of our activities and 2. Optimization. I would say north of 90% of the activation we have today uses automated strategies across platforms – it uses AI and Machine learning, So the outcomes we generate directly depend on what we measure. So, if you’re missing out in some cases 30-45% – for us it is about ensuring that as stakeholders understand the importance beyond measurement. But also the direct impact of the performance of the contents. It is crucial.   

Akash: The main challenge is that marketers will find it difficult to target the user in a holistic way.  So how would you change your thinking in a way to collect the FPD on your platform and if the user is willing to give that data to you and at what stage? As a business you always optimize for a funnel conversion and let’s say in the funnel there is a phone number and the email ID – it can backfire or elevate your life. But the point here is when and how you need to collect the FPD is challenge. Even though the CAPI as a solution sends the data to the system, again its based on the probabilistic model. I think these are the challenges, apart from that you can always look at it from a positive perspective – How to innovate with what you have. On the measurement point, how do you actually read data without one-to-one comparisons. You cannot look at campaign-level data and say that this is working. The numbers look odd and you will not be able to judge whether to put in or to pull out a budget. So how do you overcome this problem? For us we have created our own identifier that tells us from MMP that this is the source of install and from this source we track the users activity and we actually track how many users are actually converting over a period of time. To tackle this problem of measurement, we decided to guesstimate and understand the campaign performance. How the increase and decrease of performance will help you build the correlations. 

 Gaurav: One thing that we need to do better as an industry that goes beyond the negative impact and the recovery, I get this question from every single client. There needs to be better ways to get the measurement, our understanding across different advertisers that we have seen, for App there is less recovery. For the web it’s much more powerful. For Web we can recover nearly all of the signals for brands over time. So when we use FPD and cookies the entire recovery has happened over time. This is why we tell advertisers to start early because it takes time to have a first party cookie base – after 20 visits someone – for web we have recovered most of it for app there are not many offerings. A lot of advertisers don’t understand that this is a defensive technology, think of FPD as an inflation hedge – just like the wide rampant inflation we are seeing right now, you get less for your money now. FPD allows you to recover from it. You have to expect a lower baseline and you will see FPD leading to an improvement to another baseline. 

 Hicham: I would urge anyone from the measurement perspective to go through the FPD route as GC mentioned you improve mostly on efficiency. it’s not there there’s a fear but its that you can act upon it and by implementing the FPD approach you improve the efficiency. 

 How can the industry do better – What is the Way Forward… 

Gaurav: The whole polarized mindset of extremes has been a concern, what we need now is a balance and this could be like the Buddhist way – Privacy Vs Personalization. Legal theory of laws Vs the reality of marketing Vs Actual Analysis – or even wider big brother surveillance – all data is being collected Vs a total anarchy where there is no understanding of the world in terms of data. That lack of avoiding polarization is the future in my mind.  

Akash: You can keep on innovating. You will not be as efficient as you used to be, it’s a recovery mechanism. There is a 40% data loss and if this is affecting 100% of my campaign efficiency that’s the marketing money saved. Yes, you do need to invest in FPD, keep innovating how to collect this data. Funnel conversions can get impacted. We are at a stage where we have to not only rely on one way of looking at data but also need to change how data is looked and interpreted. Marketing Media Mixed Model – upcoming approach and educating approach and your counterparts and how to utilize the model of using the data on a day to day basis. Who will utilize it and create your strategy and use it to optimize at a granular level. 

 Hicham: We need to prepare where we are today across activation, measurement, Martech, and data architecture in general, and we have to adopt a privacy first measurement Infrastrucutre. There are solid technologies out there. Also to take FPD seriously, and its imp to look at how do you collect FPD – it’s also about building relationships with your customers. I think every business should do that. In terms of the future, machine learning and modelling will have a big impact for the future. You asked about B2B Businesses – they have an advantage. It’s something they’ve always taken seriously, they are not starting to collect FPD. Some verticals have an advantage over others. I would say that the answer would come from the ecosystem. I think every stakeholder would have to work together to have the privacy first infrastructure.   






Post a Comment