Marketing Analytics Trends in 2021

Dave is a data practitioner at the Philadelphia Microsoft Technology Center. He builds transformational products for customers using data. In this article Dave will look at what's hot in marketing analytics and what you should be thinking about moving forward.

Here are some marketing analytics trends we are seeing during our Microsoft Technology Center engagements in 2021.

Third Party Cookie Phaseout

Google is planning to phase out support for 3rd party cookies in Chrome sometime in 2022. Google announced this over a year ago (14 January 2020) and has already implemented some of their plan. Third party cookies are kinda creepy and many companies have phased them out already. How does a 3rd party cookie work? If you browse for something innocuous on your phone you'll see ads for that product for weeks on almost every website you visit, regardless of device. The reasons for the phase-out are clear...users want more privacy, transparency, and control over how their data is used. Chrome isn't even the first to phase these out, but since Chrome has the majority market share, the advertising industry noticed.

Unintended side effects of the phase-out have led to new tracking mechanisms like fingerprinting, which might even be worse for privacy. You can think of fingerprinting as "AI-based cookies". Companies build a profile about a consumer based on gathering your data and web-surfing habits often without asking for your permission. How does it work? Websites have a legitimate need to understand the capabilities of your browser to ensure they can optimize the user experience. So, browsers will report back information about your computing environment: browser make and version, installed fonts and plugins, OS, language, screen resolution, CPU and memory, and of course, IP address. The attributes are too numerous to list, all of which become your fingerprint. Even without your IP address you can probably be identified using fingerprinting. See for yourself.

How is a marketer affected by all of this in 2021? Quite simply, you are going to need to be more "data-driven" in your marketing efforts.

• If you leverage 3rd party cookies, you need to research alternatives. First party cookies (those under your control) are safe (for now) so you might consider leveraging those. A lot of interesting analytics and user behaviors can be culled from this data. Get creative and leverage this data in your marketing campaigns. (I can give you some ideas).
• Data management providers are trying to create industry standard audience contextual mappings that can be leveraged, safely, to create audience segments. Some of this is based on machine learning. The Chromium Privacy Sandbox is another approach gaining traction. Beefing-up your 1st party cookies using some of these tricks should provide lift. If you understand how your competition is using this data, it may help your efforts.
• You can start leaning more heavily on Google Ads (Adwords) and Microsoft Advertising (Bing Ads), although this may NOT be a good idea. Read on for some interesting stories.

Regardless of what you decide, now is a great time to understand alternatives that can provide you with a competitive advantage.

Be more purposeful with your data

Now is the time to review how you are collecting data for your marketing efforts. You'll likely need to ingest other data sets to achieve a holistic view of your customer. Much of this data you probably already have in other internal systems, it's just not easily query-able in its current state and location. You'll need to break down the data siloes and get the data into your "single view of the customer" (generally, the data lake) so marketing can leverage it.

Don't forget about 3rd party data sets available via data sharing or purchase that can provide lift to your marketing efforts. Just because 3rd party cookies are being phased-out doesn't mean 3rd party data isn't viable for your marketing efforts. Historically, 3rd party data is difficult to source and integrate with 1st party data. Data gets stale quickly, you want to leverage it as fast as you can. There are tools and techniques that can solve these problems with a faster time-to-market.

Evaluate your marketing attribution models

Half the money I spend on advertising is wasted; the trouble is I don't know which half."

--John Wanamaker, marketing pioneer

Where are your customers (and revenue) coming from? Do you measure it today? Can you measure it? When customers are inundated with your different ad campaigns how do you attribute where the customer and revenue ultimately came from? Companies should review their multi-touch attribution models to be sure the best campaigns are being leveraged. This likely means collecting new data sources with new analytics techniques that can improve your ROAS (return on ad spend).

In the past many companies just threw money at online marketing tools, often with no quantifiable benefit. Let's look at a few:

1. Uber apparently spent 2/3 of its $100 million ad budget and saw almost no change in user engagement. Why? Apparently some of that ad spend was on blatant scams. Was somebody asleep at the wheel? Real-time analysis might have spotted this earlier.
2. P&G cut $200 million from its digital ad budget, and it increased its reach by 10%!!!. Why? Theories attribute some of the spurious "traffic" to botnets. Once they analyzed their data, they could focus their efforts.
3. eBay learned these lessons way back in 2012 when it turned off its paid search ads and saw no decrease in sales. They actually did their research in a manner that would make any data scientist proud...using A/B testing...and proved that they could get the same clickthrough from search engines without paying for any search ads! Their research led to changes in ad spending behaviors by the entire industry. Good for them!!
4. Some small businesses turned off their FAN (Facebook Audience Network) ads and saw no discernible change in sales (but a huge cost savings). It's not entirely clear why this is without proper testing/analysis, but a simple hypothesis might be that the ads were targeting geographies where these small businesses don't have a presence. The point is...sometimes FAN ads don't work. Could data analytics, the scientific method, and deductive reasoning prove my hypothesis?
5. Others have reported cases of "bouncing" leading to lots of clicks (and spend) but no sales. "Bouncing" happens, for instance, when an Android user accidentally clicks an ad (I do this all the time) in an app and is taken to a website they didn't want to go too. After a few seconds the user clicks away. The solution is to often disallow ads that run on "search partners" and "display networks". A/B testing can prove if this is the case. BTW, a lot of research has been done to determine "ad placement" on mobile devices to ENSURE bouncing will occur. Apps that rely solely on ad revenue do this.

In the past marketers prided themselves on spending a lot of money on ad spend and having a presence on every ad network. That behavior is shifting, especially after 2020. Now marketers want to know if the ad spend is providing lift. That's difficult to do without understanding where you advertise and how you can quantifiably measure where each sales dollar should be attributed. Returning to this post's theme: you need to be more data-driven. Don't rely on gut instinct...it might be wrong...see the examples above.

Pandemic disruption challenges

2021 will mean a "restart" for many businesses that were disrupted by Covid-19 as things begin to return to The New Normal. Determining what The New Normal is and how it affects your marketing efforts should be top-of-mind. That might mean being asked to be more effective with less budget. But it may also mean that now is a good time to restart initiatives that were put on hold in 2020. Regardless, your efforts will need to be quantifiable. Do you have the data to support these decisions? You'll also want to move as quickly as possible on any initiatives. At the MTC, we can help you with the technology implementation. Data-driven marketing projects have traditionally been expensive capital projects with high fail rates. That is no longer the case.

Increasingly, it is the marketing department and CMO who are taking the lead in transforming companies into data-driven, analytical organizations.

Changing Regulatory Environments

Government rules will continue to change and affect your marketing efforts. In the US, privacy initiatives are being driven at the state level. In 2020 the California Consumer Privacy Act and California Privacy Rights Act set strict rules that give consumers more control over their personal information. But it also created a government agency solely focused on implementing and enforcing new rules without the need for legislation.

Other states like Washington, Utah, and Virginia are drafting their own rules applicable just to their citizens. You are going to need data to analyze how this disjoint regulatory environment will affect your marketing efforts. You need to start collecting data now.

Become Data-Driven with the Microsoft Technology Center in 2021

I am a Microsoft Technology Center (MTC) Architect focused on data solutions. The MTCs strive to be Trusted Advisors for our customers. We want to understand your business problems before rushing to solutions. We also know that technology alone cannot solve these challenges without smart people and modern processes that work. We offer services ranging from human-centered Design Thinking Workshops to hackathons.

Others have Know-How, we have the "Know-What" too.

2021's marketing challenges will require you to be more data-driven. Can you quickly do analytics on your customer data if it is siloed in multiple systems? Can you ingest new data sources as quickly as possible and then integrate them into your analytics? Ingesting data is not the cumbersome, lengthy process it was a few years ago. We use modern data tools and repeatable processes that will get the data into a query-able state in just a few days, not weeks.

We can help you build rapid prototypes for things like multi-touch attribution models in as little as 3 days. The solutions won't be perfect but we want you to walk away from our engagements excited and thinking, "yeah, we can do this!" Then we'll help you build a plan to deliver. And we're always available during your journey.

We are technology enablers, not marketing experts. We'll show you what it takes to start a data-driven marketing initiative and we'll help you solve data problems without the high project fail-rate of the past.

Does that sound compelling? Contact me or your Microsoft account team to get started.