B2B content has an attention problem - but the solution isn't real-time AI, it's subscriptions

By Jon Reed

April 13, 2022

Audio mode

B2B content marketing has a new problem, folks: we might be able to earn attention - but can we keep it? Content marketing works because buying attention doesn't.

But earning attention with content is an organizational discipline - and therefore, not appealing to the shiny new toy crowd. Most B2B companies still make the mistake of thinking that content marketing is the purview of marketers.

In reality, content marketing only works if the entire organization commits to it. If you're not an expert in creating the content, you're feeding that content to a creative team. If you're not a subject matter expert yet, no worries - you still have relevant customer-facing information. That's why marketers need a journalistic sensibility. Most B2B companies have ignored my exhortations. Then they tell me why content marketing isn't working. You can't have it both ways.

Yet there is change afoot. Look no further than Celonis, the rising process mining player that hired enterprise media titan Larry Dignan, formerly of ZDNet, to Editor in Chief of Celonis Media. Think Celonis will struggle with content much? I don't think so. They're acting like a publisher.

And yet, even with savvy hires on board, all B2B content faces a challenge. The challenge is: attention. There are two modern problems with attention: earning it, and keeping it.

B2B content competes on relevance, not entertainment

I've written about how B2B content makes its first mistake by competing on entertainment value alone, when it should compete by different standards: relevance and/or helpfulness (e.g. solving customer problems). But I must concede: even B2B content that achieves that quality threshold now runs into a higher bar: algorithmic ranking.

Increasingly, we consume content via shares on social media megasites, LinkedIn being the most important for reaching B2B audiences at this time. However, B2B content is not exactly given love by these algorithms. I'm seeing the organic reach of LinkedIn pages suffer a similar fate to Facebook business pages. Why would these sites promote page content, when their real game is advertising revenue? Facebook told businesses that our page content reach diminished because it didn't engage users. What a load of bunk. Follow the ad revenue trail instead. Here's the harsh truth:

There is no such thing as a content meritocracy. Great content is not equitably distributed. Especially with niche areas like (most) B2B content, the algorithms that dictate visibility are working against us.

B2B content attention problems - the walk-through

So, let's reframe the B2B content attention problem. Start with phase 1: Earning attention.

1. Does your content solve a customer problem, or reinforce your topic authority? (The aforementioned quality threshold, which most B2B content does not achieve) If yes, then:

2. Does your content have a distribution channel - or a chance to get shared heavily - to earn algorithmic ranking and SEO/search clout? Examples include: an author with a large LinkedIn following, or a web site blog with a high backlinks rating (SEO gold).

Let's assume your content passes the first two hurdles (good job!). Now there is another one:

The immersion-versus-interruption problem. In today's world, even when we are immersed in relevant content, and relatively focused on it, be it a webinar, blog post, or LinkedIn update, we are vulnerable to interruptions via mobile pings. Even if we enjoy the content, we might get swept up by doctor calls, Slack messaging, needy clients... We might not come back to that content.  Why? Few of us have an organized means of saving it to consume later. We leave the page (or app) - and we might not come back.

Yes, we like your content, but, there is no bookmarking of web pages anymore (well, I do it via Pocket and Evernote, but I live in outlier land). Social feeds cycle out quickly.

So, we have problem #3:

3. Earning attention is the easier part. Keeping attention is the hard part now.

That leads us to a fascinating fork in the road, one which I have debated frequently on this site. The problem: how do you keep attention with content?

Here's the emerging (trendy) solution:

Throw technology at the problem, specifically "AI." Use the power of real-time personalization to provide the right piece of content to the right consumer at the right time - in their context.

Heard that shiny new toy pitch before? Now, I'm not opposed to employing AI and automated solutions as part of content delivery (see my recent piece on AI for video editing). But: I've been burned by a so-called "AI-powered" approach to email-based content delivery that most certainly did not provide readers "what they wanted, when they wanted."

I have two problems with this solution. AI for real-time, personalized content delivery is immature at this time. Oh, and if it has any chance of working, it must be powered by unsiloed, integrated data on your audience - good luck. See: The myth of hyper-personalization - algorithms are still undermining the customer experience.

Real-time personalization via recommended content, as pioneered by Amazon and Netflix, is different. Why? Recommendation engines embedded in browsing activities are a much more feasible use of AI for content engagement. Notice how Amazon and Netflix don't push you to one piece of content: they give you a range of choices, without interrupting your browsing or consumption. Most B2B marketers stink at that; they love to use "AI" to interrupt without relevance - look no further than the invasive "what are you here for?" chatbots on most vendor web sites.

If "real-time, personalized content powered by AI" isn't the answer, what is?

So what's the alternative? Subscriptions. When I opt-in to your content, you don't have to waste money guessing my real-time context. You'll notify me when new content is available. When my context is aligned with your content, I'll check it out. And if I miss it, I've invited you to ping me again next time.

I'm not alone on this one - Barb Mosher Zinck, who knows a thing or two about content strategy, weighed in on this, via her review of the state of content marketing. Of course, subscriptions are not a magical solution. As Barb reported, many brands are wary of them:

Only a third (34%) is creating content to build a subscriber audience. But isn’t content for brand awareness, educational content, and content that builds credibility and trust, the type of content you would create to build a subscriber audience?

1. Subscription content has a higher quality bar than your corporate blog - and you must maintain the pace, or get unsubscribed. Yes. some subscription content thrives on curation, but there is an art to curation as well (Azeem Azhar's Exponential View is a definitive example).

2. Not all "subscriptions" are created equal. Twitter and LinkedIn are mostly about "following." A big follower audience is useful, but it's not as potent as subscribers. However, YouTube has downgraded "subscribe" in its algorithm, to the point that it's really just a "follow." How do I define the difference? Subscribers are notified (pinged) about new content, followers are not. Podcast "subscriptions" are in the middle - it depends on your app.

3. Email-based subscription still puts butts in seats, but it's not foolproof. Email filtering and deliverability issues can block your path, even to willing inboxes. Also, there is an emerging generation that lives out of mobile text and IM, not email. That's why you see more browser-based "alert" options. Those alerts are not as powerful as subscriptions, but still qualify as a way to sustain attention.

4. Some of the best subscription platforms aren't open to brands yet. Example: LinkedIn newsletters, a very effective content distribution channel, are only open to individual creators at this time - another reason why empowering your experts to create is a big piece of the puzzle (Though creators don't have access to subscriber information on LinkedIn, and this is often the case on third party subscription platforms).

5. The subscription "long tail" has diminishing power. It's rare for people to scroll into older newsletters. My enterprise podcasts like my Busting the Omnichannel series, which I've released since 2008, have seen an 80 percent reduction in long-tail consumption - meaning that without new episodes, the audience goes away now. YouTube is the same, except in the bizarre instances when their algorithm starts surfacing an old video in people's feeds.

6. It's a beast to personalize subscriptions to the right pace - and topics - for each reader. Elaborate preference forms are difficult to manage. I actually think "AI" can play a role here. Example: I see no reason why AI couldn't prompt an account exec to send another piece of content out, and recommend content options (this is happening already, of course). Or, alternately, AI could alert not to send out content yet, and automatically change the delivery frequency, on a per-subscriber basis. Why? Because AI isn't overreaching here, trying to guess what we are thinking in the moment, and attempting to serve up content in that moment. Instead, it's detecting consumption patterns over time, and adjusting over time - and that's right in AI's wheelhouse.

7. Audience trust is fragile. Abusing their data, especially with third party distribution (hear me, event planners?) leads to brand disillusion and unsubscribes.

My take - AI helps with content distribution, but only to a point

We are back to a fundamental problem with personalization: some claim that AI is the solution to personalization by serving up "the right content at the right time," whereas I believe the solution is guided configuration (which could be AI-powered, by the way).

Consider this botch job from Spotify, which likely has more "AI" resources than most:

(Screen shot of my Spotify playlist)

I don't like a single one of these songs - even though Spotify knows the genre of this playlist due to listening patterns. However, at least Spotify doesn't overreach and insert unwanted songs in my playlist. Inserting songs in my playlist is the equivalent of the spray-and-pray B2B spam emails that "hyper-personalized AI" systems blast out these days. The playlist Spotify's AI is trying in vain to add onto was built by super-user configuration, more than 1,000 of my favorite songs (well, at least in one genre), built over years.

Why do I believe in guided configuration? Because it is the best approximation of super-user configuration that would work for a larger (mobile) audience.

The most personalized content delivery I have is via my paid version of the Newsblur newsreader, where, over the years, I have configured and organized more than 1,000 RSS feeds. (The best enterprise content from these feeds ultimately winds up in my Enterprise hits and misses column on diginomica).

But the trend in web and mobile app design is to reduce user configuration to make apps run "simple." Netflix is a classic example: they obliterated my painstakingly configured "My List" of queued options, in favor of a wildly inaccurate "AI-based" order.

Though I was irritated by Netflix - in particular, the techno-arrogance to think they could arrange my list better than I could - I still understand their decision. As much as I believe super-user configuration is the superior approach to personalization at this point in time, I recognize that few individuals, outside of professional curators, analysts etc., would take the time I have with my RSS reader. RSS readers are like lawns; they require constant tending. That's not what our lives are about right now.

Netflix has correctly identified the problem - but the solution is not the brute force "let AI do it!" Netflix has adopted. I believe it's a guided process. I could envision a "bot" walking a subscriber through a series of options, e.g. "what topics are you interested in," frequency desired, text or email delivery, and so on, and then pre-populating consumption preferences based on that information. The back end system to distribute that content could also be enabled.

No, we're not there yet on guided configuration. In the meantime, I still believe that subscriptions - for example focused on industry verticals - are an indispensable way to get us closer to the sustained attention our content needs, until we figure out a more profound way to personalize.

End note: another way of framing this B2B content debate is between "building opt-in communities" versus "AI for real-time context."  I see subscription content as a key to building opt-in audiences. Engaged audiences can be cultivated into communities. My favorite foil in this debate is the brilliant Mathew Sweezey, who makes a highly effective case for context, and has graciously joined me in debates on this, even though we don't necessarily disagree as much as I once thought. This piece gives a flavor: From content to experience to context - a buzzword debate that actually matters. The highlight was our hour-long video debate, which remains my favorite video interview. You can catch that replay here: Debating fresh data on hyper-personalization, CX, and content strategy - with Mathew Sweezey.

And no, I'm not even touching AI as a B2B content creator. Read my dissection: Can AI displace content creators? For B2B content, the answer is no - but with a disconcerting asterisk.