Post Intelligent tracking prevention (ITP); Here are 4 things you can do to improve your tracking

Intelligent Tracking Prevention (ITP) was first introduced by Apple in 2017. ITP generally targets to make it harder to use ad tracking harshly on users, which from a user perspective is… Nice.

So far there have been five releases of ITP: 1.0, 1.1, 2.0, 2.1 and 2.2. For a full walkthrough of each ITP version, Clearcode has made a very thorough investigation.

ITP did not really, though it probably should have, set the tech marketer scene on fire before their recent ITP 2.1 update in February.

Some marketers are now panicking and think technical and tracking-based marketing is dead. Others have ideas on where to go from here.

The fact is that ITP brings along a bunch of challenges for advertisers, publishers, and tech vendors. The good news is that there are ways to deal with this, that even have potential performance upsides.

In this post we will talk about three things:

  1. What are the specifics of ITP?

  2. Which measures to take to be ready?

  3. What are the learnings and consequences of ITP?

What are the specifics of ITP?

First of all, ITP applies to Apple-owned iOS and browsers. Per the latest stats Safari-browsers makes up around 15% of browsers, while Google’s counterpart Chrome sits on approximately 63% of the market.

For the browsers and OS that are involved ie. Safari on mobile, desktop and tablet, the specific changes of ITP 2.1 means; Safari will delete all none-server side cookies after 7 days. Defacto this means the most commonly used ad cookies like Facebook, Adwords, Linkedin and so forth.

According to Digiday Apple’s ITP 2.2 update in April, further escalated cutting cookie time from 7 days to 24 hours.

Additionally, Mozilla Firefox have equally stated that they will be taking ITP steps in the future by also blocking 3rd party cookies expanding the scope of this challenge from 15% of browsers (Safari) to another at least 5% (Firefox users is a bit insecure as numbers found state everything from 5 to 15% of the browser usage market).

This means with the old ad technology of Google and Facebook where ID’s are stored in cookies, things like retargeting and attribution will not be possible beyond 24 hours. Ouch. Your move, Google and Facebook.

ITP 2.2 also forces a new user ID to be created every time you go from one site that you own and to another one. This was previously possible to work around with a link decoration of the browser cookie. For businesses using multiple domains, subdomains and landing pages, this is a major problem using a generic tracking setup.

A way to think about ITP is a cat and mouse game between Apple and Google and Facebook Ads. Apple will with each update try to close a gap. Google and Facebook ads will try to innovate new ways to retarget.

Skærmbillede 2019-06-03 kl. 12.29.44.png
Intelligent tracking prevention.jpg

4 things you can do to get your tracking ready for the future

Okay, so you see, things are changing. Surely Google and Facebook will try to fix some of this. But user protection is on the rise and will continue to challenge tracking heavy products.

The good news is that there are tangible actions you can take to deal with both current challenges as well as prepare for future ones.

  • Own your own customer data and tracking

    • First of all, you need to own your own tracking data. You do this by setting up a customer data infrastructure platform.

    • This means moving away from relying on ad platforms and Google Analytics as the main source of information for understanding your performance data.

  • Put ID in cookie and in local storage

    • Currently, most systems place an ID in the browser cookie. After ITP 2.2 this browser cookie is deleted after 24 hours. This means you have a problem if your buying cycle is longer than 24 hours.

    • For people who use Google Analytics the easiest fix for this is to instruct Google Analytics to place an ID both in the cooke as well as in the users local storage. This does not happen custom, but it’s a setting you most likely can handle on your own.

    • A system like Segment does this out of the box.

  • Identify earlier and more often

    • Positively identifying your users (ie. with mail, phone number etc.) earlier and often will become continuously more important, as tracking gets harder.

    • Practical examples of this could be to consider; gating more content behind email submission;  prompt users to login in to your product where it’s possible, basically going for micro-conversions earlier and often.

  • Do less cross domains - Stay on one domain

    • Post ITP a new user ID is created every time you go from one sub-domain / domain to another. This makes tracking significantly harder. An approach to proactively deal with this is to skip making many sub.yourdomain.com to centralising landing pages on yourdomain.com/landingpage.

    • Should you succeed in identifying the user, you need to make sure that the ID is forwarded from one domain to another.

What are the learnings and consequences of ITP?

As we see it, there’s a bunch of learnings and consequences from this:

  • Until you have fixed the ITP tracking gap, do know that the numbers/data you’re looking at in Google Analytics will not account correctly for Safari data.

  • In general, owning your own historic customer data by using customer data infrastructure platforms (like Segment) becomes increasingly important.

  • If you have set up a customer data infrastructure platform, you will not have too many attribution challenges, except for when people click across domains you own. If you do, please reach out to us.

  • However, if you are used to looking at tracking data inside Facebook and Google Ads, you will start to see they struggle to attribute value post ITP. View through tracking is not an option anymore, and click through tracking across multiple domains you own becomes harder.

  • 3rd party cookies ie. Google and Facebook ads can no longer retarget beyond 24 hours (the time of which Safari will delete the cookie). This means one of the most simple marketing setups “buy traffic and retarget” is taken out of the playbook.

  • This logically also means that ad platforms will try to go from being a 3rd party cookie to a 1st party cookie.

  • DMP’s that have heavily been relying on sharing 3rd party data will have to find a new way to operate.

  • 100s or even 1000s of landing pages on sub- or independent domains should probably be taken out of the playbook, as tracking a user is no longer possible.

  • On a positive note, websites will become faster as you are not sending a stuffed cookie back and forward every time the user clicks another link on your website. As most internet marketers know: Better website performance equals higher conversion rate.

Dreamdata will continue to follow the ITP development and share our take on it with you.

If you have things to add or suggest to the post or if you have found good methods for growth beyond ITP, we would be happy to hear from you.

How to do attribution with Segment

Understanding what activities and campaigns contribute to your company’s revenue and what, in lack of better words, is a waste of money, is critical for a business to know.

At Dreamdata.io we’re on a mission to build the world's best multi-touch attribution tool for multi-techstack businesses.

We’ll probably never get to a 100%, but what we are able to tell you is, in our humble opinion, as close to the truth as anybody can get you.

We need three ingredients to do so:

  1. Your traffic drivers and the software tools you use

  2. Information about your users behavior related to your business and your website

  3. Data from your CRM and revenue software where you register your deals and revenue

When we have all three ingredients the Dreamdata algorithm can then stitch the multitude of data together, remove redundancy and present an easily understandable, non-biased models for analysis and decision making.

This further allow you to integrate your customer data ecosystem including rich tracking data for true behavior analysis and heuristic and algorithmic attribution models that handle multiple touches across many users.

Traffic drivers and software tools

Despite ever more tools and data available, the B2B buyer's journey is becoming ever more complex. Multiple people have a say. Multiple platforms, tools, media and touch points are used to impact decisions.

Investors are shouting scale. Diminishing returns haunt your marketing spend. CAC vs. CLV is moving in the wrong direction. Sounds familiar?

What actually drives your revenue?

Your CMO will give you one answer. Your sales guy another. And support will claim their stake as well. Oh yeah, and the product people are convinced it’s their new feature that works.

All of them are probably partly right and partly wrong.

What you need to do go through your whole commercial organization and ask everybody which tools and software they use on a day to day basis.

  • Where do marketing find traction, leads and automate mails from? Ie. Facebook and Google Ads, content marketing and search, activecampaign, campaign monitor and so forth.

  • Where do the sales department log revenue, activities, pipeline, calls etc.? Ie. Salesforce, Hubspot and Pipedrive.

  • What tools do your support and customer success teams use to provide your customers with stellar help? Ie. Zendesk, Intercom etc.

I think you get the point. Everybody contributes to revenue. Not just one department. Each contribution needs fair attribution or you end up making wrong decisions.

Wherever your customers have a touch point with your business you need to recognize it, note it down, forward the data to a database platform and have their fair share attributed to what made revenue happen.

You can either gather all this data in warehouse solution like Google Big Query or use Dreamdata.io.

Attribution to segment.png

Understanding your users behavior with Segment

One of the key ingredients in this attribution soup is how your customers are behaving from when they come to your website the very first time and all the time and touch points on their user journey until they end up buying your product or service.

Back in the day, you would think Google Analytics was enough for this, but the truth is that it comes up very short in today's very cluttered data and tool space

To get clarity on your users behaviour it’s very recommendable with a proper customer data infrastructure platform.

At Dreamdata, one of the partners, who we regular recommend companies who are serious about growth, to use Segment.com as their customer data infrastructure.

With Segment you plugin all your customer data and clean, collect, and control it.

This means you can integrate all of your favorite tools and move their data to where it’s needed ie. send it to Google Big Query (a data warehouse) as one part of building your own analysis.

We recommend Segment because it’s state of the art for this and secondly, because we have already built an integration that with one click allows you to send your data through for further analysis.

Segment to big query.png

If you are already using Segment, but are struggling with a few things, we are here to help you iron out the last details. We are both Segment Certified partners and have been using it for years and seen most bugs, so in that respect, don’t hesitate to reach out.

Connecting the dots from anonymous user id to revenue in your CRM

The last piece of the puzzle you need before you can start to do proper attribution is to understand how much revenue the person who came in as an anonymous ID ended up providing as a known customer of yours.

You do this by connecting data about raw ad and campaign performance with user behavior and revenue from the first purchase to life time value (LTV) contributed in your data warehouse. This is particularly important if you are a B2B Saas with target CAC set based on LTV.

Note: For this to work well, the importance of your sales team being disciplined about logging activities and representing true revenue numbers can not be emphasized enough.

Segment plus CRM.png

When you have identified and gathered, cleaned and connected all your data inside the data warehouse, you are ready to build your own attribution models and analysis.

For most companies, it does not make sense to take a complex project like this on. We did it. It was hard, but we made it work.

We now know how to do it and that's why we decided to built Dreamdata to save you the hassle  and complexity. Instead you can jump straight into easily understandable, non-biased attribution models for analysis and decision making. What’s not to like?

Where do you go from here?

Having set all this setup and connected you will need to strategically consider what attribution model(s) that you want to strategically drive your business after.

We will leave this for now and return to the topic in another post.

If you need help getting attribution right using Segment, please do reach out. We are experts and can help you become so as well.