Content, Campaigns, and Context: Why CDPs Are Having a Moment
Someone tweeted about me the other day:
@gadgetopia loves to hear himself write.
And it’s true. I love to write. And I write for myself, really.
If someone reads me, that’s awesome, but it’s not really required. Honestly, I’ll just yell into the void all day long.
I’m content-centric. For me, creating the content is the point. I need to do it at some base psychological level, and I guess it’s just a bonus if it coincides with a business goal.
The average organization has historically been content-centric as well. We create content first, and we hope that it has some value to some person down the road. The value is usually just assumed, because you can’t exactly predict these things, but we hope we do okay.
Personally, my feeling is that if you create good content, it’ll pay off at some point. Call that “shotgun content strategy” – shoot a lot in one general direction, and eventually you’ll hit something. This is grossly inefficient, but a hard thing to accept is that creating content is the easiest part – getting an actual return on that content is harder.
(Note: I showed this post to my marketing manager. She said, “When I was at an agency, we used to call that ‘spray and pray’ marketing.”)
Exhibit A: the content management system. We build our processes around systems for creating, editing, and managing content. Our strategy often starts and ends with content – we’ll create it then throw it out into the world and hope that it works. There are people out there consuming it…somewhere. We hope they like it.
The dream of content-centrism is something like throwing bread on the ground at the beach, and watching swarms of seagulls descend on it and fight over it. Wouldn’t this be great? We don’t know anything about the seagulls or their story or their relationship with the bread. We just toss it out, and they descend, and we hope this is how it works with our content.
This is First Wave Content Marketing: content management systems as the foundational pillar of a content marketing infrastructure. Names like Episerver and Sitecore and Drupal loom large here.
The next stage of maturation is the Rubicon an organization crosses when they go from content-centrism to campaign-centrism. There’s always a very different vibe to organizations that speak of “content” and those that speak of “campaigns.”
The former will say “we published a blog post!” and the latter will say, “we’re seeing great conversions and engagement from our last campaign!” The former just throws content into the void (guilty), while the latter plans coordinated marketing campaigns around ideas they what to communicate, then builds content in support of those ideas. They can’t imagine just creating content without being part of a larger effort to influence.
The campaign is a conceptual framework of positioning and framing. The content is just what gets bolted onto that framework. The campaign became The Thing – content was now just created and managed in support of The Thing.
This shift changed how we sell CMS. We used to lead pitches with management – “This is how you edit a page of content.” But then the market changed, and we started to lead pitches with marketing – “This is how you set-up an A/B test campaign.” It was a tectonic shift in how this industry was structured. Management became secondary to marketing, and content became secondary to campaigns. This was a healthy change.
So, let’s call this Second Wave Content Marketing: when campaign management began to take precedence over content management. This is when names like Pardot and Eloqua and Marketo started to creep into more and more conversations. In addition to a CMS, people wanted to know how that was going to integrate with their larger marketing platforms.
I feel like there’s a Third Wave of Content Marketing coming, driven by a technology shift called Customer Data Platforms, or CDPs. This is where content and campaigns become subordinate to context.
Haven’t heard of these? You will. (apologies to Tom Selleck)
[…] packaged software which creates a persistent, unified customer database that is accessible to other systems. Data is pulled from multiple sources, cleaned and combined to create a single customer profile. This structured data is then made available to other marketing systems.
The idea is that your CDP is the single source of truth for your customer data and a history of their behavior – their “context” with you. A CDP tracks customer activity across every channel, even when the customer can’t be identified.
If your customer identifies themselves at some point (ex: they register for a webinar), their profile is “de-anonymized” and can then be matched up to all sorts of other information, and you can begin to push known information about that customer into your CDP.
(And, yes, this is where some vaguely creepy things can start to happen with data purchased from profiling firms, but that has nothing to do with a CDP, directly. A CDP is just where organizations put that data, alongside lots of other – less creepy – data.)
Two differentiating points to note:
This is not a CRM. A CRM tracks the actions of humans dealing with humans around specific tasks – selling or supporting something. You’ll often push data from your CRM into your CDP, and vice-versa.
This is not analytics. Analytics is usually channel-centric. They tell you what aggregate visitors did in aggregate on a particular channel, not everything a specific person did. Additionally, analytics are historically declarative (“this thing happened”), not an aggregation platform – put another way, you don’t go pushing data into them after the fact to reference later.
A CDP is a central clearing house for massive amounts of data about your customers’ behavior. It might share data with a CRM in both directions, but it’s very much a separate segment. A CDP gives you intricate details about their behaviors – “Person X took Action Y in channel Z” – and gives you a place to put data about them that you source externally and after their direct interaction with you.
CDPs are suddenly a big deal in the digital marketing world. Vendors are sprouting up everywhere. CMS vendors are offering integrations or building their own offerings. Marketing suite vendors are adding CDPs as core parts of their family of tools.
You might be thinking, “Didn’t big companies already do this?” And they did. But these systems have never really been productized – bigger, more advanced organizations just built their own, and only now are patterns emerging that are driving generalized products.
From my friend Tony Byrne from Real Story Group –
The search for unified customer and prospect data is not new. For more than a decade now, larger enterprises have been custom-building – often at great effort – what RSG calls “Customer Data Backbones.” These were typically API layers built with traditional enterprise integration toolsets and data warehouse platforms. Early-adopter enterprises have built impressive customer data backbones, but the resulting layer tended to be very bespoke and IT-intensive and therefore not as adaptive as agile marketers always needed.
CDPs are the technology industry’s attempt to replace custom backbones with more marketer- friendly, packaged solutions.
TLDR: we already did this ourselves, just like we built custom CMSs for a long time. Companies are just productizing these now.
And this is why the shift is coming: when something gets productized and ultimately commoditized, it becomes part of the common lexicon of an industry. Now we can all talk about a “CDP” and know we’re discussing the same thing – either an actual platform/tool, or the generalized concept of warehousing customer context data.
The “data warehouse” analogy is apt. These are data warehouses for customer behavior, in the sense that they’re designed to store and make sense of data pushed into them from all directions. And like that segment, custom solutions without names existed for a long time.
The adoption pattern might be similar too – back in the 90s, before the data warehouse market became productized, this was a space where every company just built their own thing.
But then someone slapped a name on it – Bill Inmon published Building the Data Warehouse in the early 90s (I read it when I was starting my career in IT a few years later) – and now there was a “named market” that vendors could coalesce around. We weren’t just building consolidated databases – we were data warehousing! Sometimes all anything needs is a name to give birth to a segment (and a million “expert consultants”).
There’s a developing OASIS standard that was spun out of Jahia, the WCM company. It was originally called a “context server,” but then they conceded that a name had happened:
Since the OASIS Context Server TC has been established, the term Customer Data Platform (CDP) has emerged and can be interchangeably be used for the Context Server. To reflect this, the specification produced by this TC has changed its name to : the Customer Data Platform specification.
I actually like the “context” name better. That’s what this is – the context that your customer has with your digital channels. If you ever ask the question, “What does this person know about us?,” then these products seek to provide that answer. They want to record the context between your customer and you.
Organizations are acquiring packaged CDPs from three main sources:
- Pure-play vendors
- CMS vendors
- Marketing suite vendors
There are slew of pure plays in this space, almost all of them SaaS vendors.
I know, I know – you haven’t heard of any of these. That’s fine. Back in 2003, you hadn’t heard of WordPress either.
RSG is currently covering 26 pure plays and another 3-4 larger suite vendors. I spent some time with a handful of reports that RSG was kind enough to provide. Axes of evaluation seem to be:
- Ingestion: how easy is it to get data into the platform?
- Analysis and Segmentation: how can you play around with the data inside the platform?
- Usability: what user level is targeted by the platform? Developers? Marketers?
- Connectors: what external services does the product integrate with, either for automatic data ingestion or more complicated analysis?
Vendors are all over the map, both in core functionality and also seemingly in targeting. Some products are for developers, some marketers, some for data scientists.
Another problem – simultaneously more mundane and more dangerous – is that no one is disclosing revenue, so you have no way of knowing who the real players are, and who’s going to be around when the inevitable shakeout and consolidation happens. That problem is what tends to drive people away from pure plays and into the arms of their existing marketing suite or CMS vendor.
So traditional CMS vendors are jumping into the fray with their own solutions. Here’s Joey Moore from Episerver about their Insight product.
The digital experience platform is the natural home of the CDP. This became increasingly apparent, and why we developed our own solution some years back. Clients were pulling customer data out of the platform to use somewhere else or pushing external data in, often from solutions that we now provide (think email marketing or product recommendations). So in that regard we kind of had a head start as we understood the lifecycle and the value of the data that was being captured.
Fundamentally a CDP comes down to three things; Scalability, there’s only ever going be an increasing amount of data. Interoperability, to realize the value of the data, you need to be able to put it to work. Security, your customers trust you to look after their data and breach of that trust can cause irreparable harm.
And here’s Mark Demeny from Sitecore:
We’d agree that we’re in the midst of a third wave of content marketing built around context, and that’s been one of the guiding principles for our development efforts for quite some time now, particularly with the xDB, where our primary focus wasn’t to provide analytics for historical analysis, but to provide a foundation for audience interaction (and be deeply integrated with our marketing automation and other tools).
In the past, there was some opinion that CDP would be subsumed as a feature into other marketing and advertising technologies, but shifts in the market have changed the dynamic towards something more capable of handling the complexity of the current state of the market.
First, the number of channels is starting to increase to include things like offline, support, voice and IoT that don’t necessarily go through the web and martech channels particularly easily. Second, first party control over how your customers choose to interact with your brand is even more important in the age of GDPR and diminishing comfort and ROI from social channels.
But, like any move into an adjacent market, not all CMS vendors are convinced this is the right direction. Here’s an opinion from a mid-tier vendor who asked to remain anonymous, as they’re still watching the market develop.
We’re all for using customer data to improve content choices, to sharpen customer experiences or to upgrade personalization. Plenty of our customers have been integrating our CMS with their custom-built CDPs in the past and we now see customers making the connections between their CMS the new productized CDPs. However, the risk of feature creep and missing clearly identifiable excellence in the CMS is palpable. We don’t want to be yet another DXP that tries to do everything but can’t really get the fundamentals right – or which cannot realize the promised benefits of the integrated suite.
The trend toward data-driven content and experience management is great – we just believe CMS users are better off handling their marketing use case needs through more specialized tools that connect well.
The marketing suite vendors are all jumping in as well. A couple of them declined to comment, but Adobe and Salesforce both announced CDP plans in the last week.
After the Adobe and Salesforce announcements of entry into the CDP space, very few people doubt its usefulness. However, many people still rightfully challenge the ability of CDPs to deliver value, as the value is only seen after integration with channels that allow both transactional and personalized communications.
I love that quote because it says, “This is something we’re all super-excited about, but we’re not sure everyone can make it work…” Which, let’s face it, is the story told of all higher-end marketing features from the last 10 years.
Regardless, in the end, all indications appear that this is What’s Next. (apologies to Martin Sheen)
In writing this, I’m not trying to savvy anyone. If you’re creating good content regularly, good for you – you’re doing better than most, even if you’re just yelling into the void.
But, the natural maturation of a content marketing engine has traditionally moved that focus from content to campaigns. Organizations that got there were once on the leading edge. But the advent of the CDP signals that we’re moving up the scale to organizations that have comprehensive, contextual customer tracking, and who develop campaigns in response to that data, and target content directly to defined segments of that data.
A gross over-simplification: Good content gets you to 90%. A good campaign strategy gets you to 99%. Perhaps understanding and leveraging context is that last 1%?
(Until, of course, the next thing comes along. To preserve my alliteration, I hope it starts with a hard “C.”)
And this brings me to my final point, and what I think might be the natural end stage of this trend –
A CDP might become the foundational pillar of a content marketing engine. To date, we’ve been fixated on CMS. But when undergoing a digital transformation (a phrase I loathe, but which fits here), organizations might shop for a CDP first, then layer a CMS over top of that. The CDP will be the core of their content efforts, not a CMS. The CMS will just serve in support of a CDP.
I’m interested to see how the market develops. The current batch of CDPs aren’t cheap, and there doesn’t seem to be much of an open-source community around them, despite Apache Unomi being a top-level project and the reference implementation for the OASIS standard.
This article at CMSWire puts some more context around Unomi – notably that it’s a “headless CDP,” to which you bring your own interface or third-parties start selling productized interfaces and data mining solutions for it. (Trivia: I can’t find confirmation of this, but the fact that Unomi sounds just like “you know me” cannot be an accident.)
I think we might see three-way a digital divide, where we group organizations by what “C” they’ve achieved:
- Content-centric: most all organizations will remain here
- Campaign-centric: will require some investment in marketing automation and considerable investment in staffing
- Context-centric: will require a further investment in CDP tools
I’m a old-school, server-side content guy. I was never comfortable with marketing technology, honestly – I believed in the management and publication of good content, full stop. Start with that and you’ll be fine.
But, weirdly, CDPs don’t freak me out. And part of this is probably because I’ve been selling the idea of human-centered content for more than a decade, and the promise of moving the locus of control from the content to the customer seems like a natural, healthy progression to me. I particularly like the separation of customer and behavior data from content – I feel like there’s a crossing of concerns there otherwise, that makes things muddier.
Also, I’ve mentioned the phrase “content marketing” a few times in this post. It’s little-known fact that when you say that three times, Rob Rose appears, just like Beetlejuice. (I’m kidding – Rob has been a friend for 10 years. I sent him the draft of this post and asked him to comment.)
Successful Marketers are starting to look at their databases as “audiences” rather than simply “leads” or “opportunities”. The pressures of GDPR and other data management challenges require companies to take a CRM-like approach to their classic marketing data. This is where CDP’s can truly serve as the pre-customer database. This is where marketers utilize (and manage) a much more holistic (and cross channel) view of audiences and their preferences in order to develop trust much earlier in their journey.
Here’s what I like about this quote:
"audiences”: We do our best work for audiences. We perform for audiences. Whether it’s theater or digital, a more concrete concept of the “audience” makes us better at what we do. Editors like to know their stuff is being consumed, and if they can have a longer view of that context, they’re likely to get higher satisfaction from content creation.
"pre-customer”: This speaks to the fact that with a CRM, you’re talking to customers or direct prospects. A CDP, rather, is elevating your content consumers to a higher level. You can still push content like crazy, but now you’re taking direct steps to realize some return on that.
"develop trust much earlier in their journey”: And this has always been the implicit dream of content marketing. We’re going to show you what we know and what we can do before we even start talking to you about a sale. Hopefully, we’ll build credibility (trust) before we unleash sales on you.
If this reasoning helps organizations push more content because they can realize a return on it more easily, then that’s a good thing.
CDPs might provide the justification and ROI necessary for organizations to invest in better content. If they can identify a piece of the ROI puzzle earlier, then more exploratory content is easier to justify. A CMO might throw more money at content creation if there’s a mechanism for them to isolate a return on it, and a CDP might be that mechanism.
I’m here for that. We all should be.
This is item #2 in a sequence of 285 items.
You can use your left/right arrow keys or swipe left/right to navigate