What Does 1st Party Data Mean To You?
dall-e 2 prompt: someone joyously recognising how much first party data they actually have

What Does 1st Party Data Mean To You?

What is 1st Party Data?

Third-party data (3PD) is data owned by someone else. First-party data (1PD) is data that you own. This definition, along with what a CPM is, why CTRs are so important (I think I might have missed this one), and the name of your local pub, are among the first things that you learn when starting out in the ad industry.

Yet this simple definition of 1PD being anything owned by the creator of it is often forgotten by media owners when building a commercial data proposition.

How did it all start?

Rewind 5+ years ago, there weren’t an abundance of publishers that offered a productised, commercial data offering. Plenty had a data management platform (DMP), propped up significantly with 3PD enrichment, but very few had a compelling and well-branded offering to take to market. (I recall in the earliest stages of our DMP usage at Immediate Media they were just referred to as “DMP deals.” Things have changed a lot since.) We’ve seen a brilliant evolution in our industry since then, however, in which the majority of large-scale publishers, certainly in the UK, have built interesting audience propositions with targeting and insight capabilities. Advertisers now have a buffet of high-quality data offerings to choose from.

every data asset they (publishers) own, whether it’s behavioral, contextual, revenue & performance, panel, trend, editorial, and beyond—all of this and more can be a data source of value to advertising clients.

What the best have done, however, is take this one step further by coming back to that simple definition of 1PD. They recognize that every data asset they own, whether it’s behavioral, contextual, revenue & performance, panel, trend, editorial, and beyond—all of this and more can be a data source of value to advertising clients. This is a step change in the traditional definition of commercial data in which we move from just audience or standard DMP functionality to something much more extensive and indeed useful to a client.

Let's expand on just a few of these areas.

Behavioural data+

Telling an interesting story about the content that is being consumed on your platforms and creating cohorts based on this is the bedrock of any data offering. But there are some straightforward ways in which you can elevate this capability to the next level. Most of us, for example, will collect declared data in various ways. Leveraging this as a segmentation tool to collect data on the areas that are clear blind spots for your business (e.g., we know at goodfood.com that lots of people shop at Sainsbury's, but identifying this group can be tricky) can dramatically enhance your offering, especially if/when you are bringing this data back into your primary segmentation tool.

You can also achieve a lot, particularly for those non-endemic briefs, by falling back on panel-based data to build a story around the data you have that plays to your strengths. If, for example, you work for a lifestyle publication and you receive a brief in which the targeting is high-income males aged 25-44 (yawn), it might be that you don’t have any demographic data beyond 3PD that you can target. Besides, anyone could offer this.

Instead, through panel-based research, you might find that this group of users tends to highly index against premium travel content and on average spend 2.5x more time looking at fitness and health content than the average. Perfect, you can create cohorts on these precise areas, meaning you’ve not only told a story to the client with 1PD but then utilised it for your targeting recommendation.

Context vs. Data – the rivalry that never existed

Data and contextual are two sides of the same coin. The best contextual solutions are now underpinned by a data-driven approach in which intelligent models process the content and can then understand its meaning. If legacy tools took a simple keyword identification and categorisation approach, evolving solutions instead take all the data associated with a piece of content, feed it through a model which then makes intelligent decisions on what the true context is. This enables the creation of novel new cohorts that are unrestricted by the words that may or may not be present on the page and instead based on data that could be associated with content or even the customer looking at it in that moment.

Your contextual capabilities should therefore be part of your data offering. When viewed in this way, you can frame a full spectrum of data-driven targeting capabilities to your clients from the highest level, broad contextual approach through to the most granular and bespoke segmentation based on a small seed data set.

Buyer signals everywhere!

We all know that the future of digital advertising is one in which there’ll be fewer IDs. I don’t necessarily agree with the perspective that more broadly signals will be in decline. I think it is just that the makeup of those signals will change. Publishers will continue to have an abundance of signals and if they’re smart, be able to in fact increase what they can provide to the right strategic partners. We are also seeing some of the largest SSPs seeking to take advantage of this opportunity.

The point being that in the case of programmatic advertising, publishers hold a huge amount of 1st party data on every transaction that takes place on their site(s). Wrangling this data can be challenging but what we choose to do with this shouldn’t just be focused on partner optimisation, price flooring, and wider business unit insights. Take, for example, the audiences which you knew in that precise moment the user fell into when an advertiser delivered an impression. The articulation of the story around what this means for an advertiser and how you could recommend improving their outcomes, recognising that they’ll be increasingly buying blind, can be a clever use of an under-utilised 1PD asset. You’ll want to create something that is scalable of course but basic dashboards on the top 10 audiences and some other visuals for any of your top 250 programmatic buyers can supplement any wider sales pitch, particularly when it’s focused on converting that to a direct relationship.

It’s time to be industrious!

If your organisation is utilising a data asset for business decision-making, the chances are it could be of value to you and your advertisers. Editorials are a great business unit to illustrate this. These teams are hugely dependent on data to understand current trends and anticipate future ones. Now, for many publishers outside of the most advanced ones, the tools and insight here aren’t utilised by the ads team. Think of the power you could wield by articulating a clear story to your advertisers around trends you have visibility on that literally shape your content strategy. A treasure trove! Identifying what these primary sources are that editorial relies on, establishing a relationship to make these accessible, and distilling this down into your wider commercial data playbook can elevate your wider offering substantially.

Think of it this way:

  • Our latest research shows attitudes to X are changing to Y (1P panel-based)
  • This is mirrored by trend X that we are seeing on our sites (Editorial trend-based)
  • Wider market intelligence also shows Y (Panel-based tool)
  • Buyers are shifting spend towards more X (e.g., 1P programmatic data)
  • We are seeing our users behave in this way on our sites (1P Audience data)
  • We also know this about them (1P Audience data)
  • And this is the best way to reach them (1P Audience & Contextual data)

This is an example of how, when 1PD is viewed through the lens of being all of the data that we own, we can craft very compelling and, crucially, unique stories for our advertisers.

I’m not advocating, by the way, for the senseless pursuit of ingesting and processing as much data as possible. We know that more data doesn’t always mean better outcomes and the comprehensive impact that data centers are having on our environment is becoming well understood too.

Look under that sofa and you might be surprised to see how many coins (1PD) you find.

My point rather is about recognising what you currently have and how this can be not only better leveraged but brought into a more cohesive and captivating proposal. Sure, you might identify some additional platform integrations or data tools to supplement your pre-existing capabilities, but when you reframe your understanding of 1PD, you’ll see that you probably have something interesting to tell any advertiser and this doesn’t have to be an extensive 10-page deck of insights. Provide them with something they couldn’t reasonably have been aware of and it’s the first step in establishing new relationships. The smallest of insights such as “X% of the inventory you’ve delivered programmatically on our sites reached this target audience” or “We can see you’re spending 2x the average to reach this type of user” or “Our latest research shows X% of the population intend to purchase Y in the next 3 months and we’ve built a custom segment on this” can go a long way.

Look under that sofa and you might be surprised to see how many coins (1PD) you find.

?

Luke Judge - hundo

CEO of hundo.xyz - An Immersive Work Preparation Platform / Non-Exec Director / StartUp Mentor / Marketing Specialist

3 个月

Great article, Matthew Rance! Especially agree with the point about leveraging behavioural and contextual insights for better targeting. Also commenting on this for more visibility across my network! Good job.

Tom McKay

Director of Client Success & Strategic Growth

3 个月

This is a really great piece Matthew Rance , while education on the quality and diversity of 1pd has to happen at a buy side level, there is still significant work to be done in-house at media owners to educate on what makes up a well rounded dataset

Alessandro De Zanche

Independent consultant - Senior advisor - Audience, advertising, media monetisation strategist

3 个月

It is refreshing and energising to read your post, Matthew, as there aren't many media brands thinking in this way (yet). It is a holistic and proactive approach, taking the initiative, building propositions that advertisers (and internal teams too!) can benefit from. It's the the first-party web mindset, as opposed to the one of using a DMP to sprinkle third party data on top of the inventory and declaring "I have a data strategy". And I think it is important to keep remembering how fundamental it is, as you have reminded too, not to decouple that data from the context and the environment, which is what ultimately provides uniqueness. Many non-media companies can offer amazing 1p data too, but it is the whole consent + data + context + engagement and attention + trust that makes premium media owners unique.

Keshav Parthasarathy

Senior Product Manager at Expedia | Ex-Ocado , Ex-McKinsey | Commonwealth scholar | LSE

3 个月

Nice article Matthew Rance. What is your view on data enrichment/data collaboration via clean rooms? The balance of getting other signals versus customer data being combined and floating across clean rooms?

Mattia Fosci

AdTech, Legal and Policy Expert | Advocating ID-less Digital Advertising | Founder & CEO @ Anonymised

3 个月

Great post Matthew Rance - it should be an interesting read for the Competition and Markets Authority

要查看或添加评论,请登录

Matthew Rance的更多文章

社区洞察

其他会员也浏览了