Tag Archives: relevance

The urgency of genuine personalization

Personalization may be the most misused phrase relating to content.  It’s not hard to understand why people want to talk about personalization: it’s appealing to think you’ll see exactly what you want, especially has we get deluged with content.  But paradoxically most techniques of personalization actually involve tailoring content based on what other people do, rather than your own interests.  As a result, people miss out on content they might most want to view.

To get a sense of why personalization is so urgent, and so troublesome, consider the situation of Facebook.

Mark Zuckerberg said last year that Facebook wants to be “the best personalized newspaper in the world.”  Notwithstanding Facebook’s popularity, its users complain about the lack of relevancy for much of the content they see.  Facebook is optimized for promotion of content sharing, not for filtering of content based on individual preferences.  Those two goals to a large extent are in conflict with each other.  Facebook has chosen to fund its revenues through advertising and related services, which  accounts for about 90% of total revenues.  Brands want their content viewed and shared and their ads seen, so keeping them happy is a huge priority.   These various pressures come to a head with Facebook’s News Feed.  On average, a person may have 1500 potential messages Facebook considers relevant to them, and Facebook needs to prioritize these into a manageable quantity (they’ve decided that’s about 300).  “The News Feed algorithm responds to signals from you” Facebook explains.  But many signals seem to have little to do with the individual, and more to do with other parties: the interests of friends, strangers, and advertisers.   Factors influencing ranking include “number of comments, who posted the story, and what type of post it is (ex: photo, video, status update, etc.)” and “promoted posts.”  Facebook routinely revises its algorithm based on what tests show people in general like best.  But the options for the individual to choose what he or she wants specifically are few.  Facebook decides what types of content, and items of content, are most relevant, and individuals don’t get much choice in the matter.

What do we mean by personalization?

In the digital content arena, personalization lacks any widely agreed definition. As Aria Haghighi of Prismatic notes: “personalization is really young. I still think we don’t all agree necessarily on what personalization means.”   Part of the lack of agreement involves how to implement personalization on a technical level, but also it reflects a lack of common vision from providers about what they want personalization to offer.

Digital marketers generally refer to personalization as behavioral targeting, and often use the terms interchangeably.  Big data researchers typically see personalization as adapting content results on the basis of machine learning.  Curiously, the protagonist of the story, the person seeking content, is missing from the personalization discussion.  Instead, the discussion is centered on how to improve click through rates.

Serving people better should be the core reason for personalization, and it’s important to build a commonsense definition, rather than a mathematical one.  Merrium Webster defines personalize as “to mark (something) in a way that shows it belongs to a particular person.”  The key elements are that it is individual to a person, that the person owns that something.  Ownership implies having control.

My definition of personalization is when a person gets unique content that reflects their individual preferences.  Targeting, in contrast, is when a person gets non-unique content based on characteristics they share with others.  In both cases, the content provider prioritizes what content is delivered, but in the first case it is based on first-hand knowledge of what an individual is interested in, where in the second case, it is based on second-hand assumptions about what seems relevant to the individual.

Personalization is based on explicit individual preferences, not assumptions

To understand personalization, we need to separate two dimensions:

  • whether the “signal” is about the individual himself, or about the interests of a crowd who are assumed to be similar to the individual
  • whether the “signal” is an explicit expression of interest, or an implicit assumptions based on prior behaviors

The following table shows how different signals can be aligned with either individual or crowd, and can be either explicit or implicit:

Table showing how different kinds of explicit and implicit preferences and behaviors can influence content delivery
Table showing how different kinds of explicit and implicit preferences and behaviors can influence content delivery

I’ve simplified the classic approaches of Facebook, Amazon, and Google to highlight elements that are most salient in their respective approaches.  In practice, each uses a mixture of signals to rank and filter content, crunching hundreds of different largely behavioral signals.  These high volume content providers, and others who are far smaller,  offer individuals the impression that the results are personalized (described as being “for you”) when they are primarily based on the aggregation of data across users, rather than individual feedback.  While these aggregation techniques improve general relevance (fewer inappropriate items), I don’t believe these behavior-driven approaches are sufficient to give individuals what’s most relevant to them personally.

How to implement personalization

Personalization matters because it is the only way individuals will be able to cope with the volume of content they face now and in the future.  Too much information is the problem, and genuine personalization needs to be the solution.  Brands need to help individuals connect with the content they most want, and not simply content that’s an adequate fit.  To do that, they need to ask questions, and not just make assumptions.

Lots of content providers talk about offering personalization, but the techniques they rely on have big weaknesses.  In future posts, I will discuss why big data approaches can’t solve personalization, and why small data using individual feedback is essential.

—Michael Andrews

Why your content is never up to date

Audiences need to feel confident that content they are viewing is active — current, accurate and relevant to the moment — and not dead.  They expect to view the newest information. A robust process for keeping content up-to-date has never been more important, but existing approaches fall short of that ideal.  In part one of this two part series, I will examine why common approaches for updating content can create problems.  In part two, “Making content updates an intelligent process,”  I will suggest alternatives.  First, let’s look at why the current situation is a problem.

Audiences can punish brands on social media when content’s dated and inaccurate.  Even content with minor blimishes, such as copy that simply looks outdated, can influence the perceived credibility of the content.  People don’t want to think they are getting something that’s old and stale, something that makes them feel they aren’t a priority.  Numerous studies have shown that audiences evaluate information for being up-to-date ahead of any other factor.   Current content is also essential for SEO.  SEO experts believe Google gives priority to most recent content, and content that appears more recent will probably have higher click through rates (CTRs).

Brands still face major challenges fighting the problem of dead content, even though everyone agrees on the value of keeping content up-to-date.  Why is this?  The old model of “publish and forget” is no longer acceptable.  Previously, brands might publish something, then mentally throw it away after it was no longer desirable or needed, without actually taking steps to retire the content.   Now, content will often have at least a notional expiry date associated with it.  Deadlines sound serious, as long as people believe in them.  But many times the content updating process is built on wishful thinking.

Common approaches to updating

The current approach to preventing dead content is what I will call “publish and be vigilant.”  This approach involves several tactics that rest on two core values: discipline, and accountability.  The vigilant approach depends on the heroism and foibles of people.  It can work well in some situations, especially when there is a small group of content creators who regularly revise content.  But a process based on vigilance is not effective or sustainable for large scale content management.  To understand some the elements of the vigilant approach, we will look at the content governance advice of the U.S. Government (as of this writing in early 2014).  I’m using this example not because it is notably good or bad, but because it is representative, and provides a rare glimpse into an internal organizational policy.

screenshot of content updating advice
Advice from the U.S. Government’s How To website (screenshot)

The above screenshot shows the basic process.  The key recommendation is to set up a review process: content owners follow a list of criteria to check that the content is readable and functional, and conduct reviews on a quarterly basis, with all content reviewed once a year.  This advice is consistent with recommendations commonly offered by content strategists, and is sound as far as it goes.  These things do need to happen, even if they may need to be done differently than the basic guidance would suggest.  But the larger problem is to assume doing these things alone will be sufficient to keep your content up-to-date.    The real world of content is far too messy to be managed by simple rules.

Hope #1: Content ownership

All content needs an owner — that’s the standard advice.  Here’s what noted content consultant Garry McGovern says: “Anything that does get published must have an identifiable owner. That owner must commit to regularly (every six months at least) checking their published content. It is absolutely no excuse for them to say they don’t have time. Don’t let them publish if they don’t have time to review and remove.”  But McGovern also notes “It’s very hard to review and remove. Not alone does it take time, it also takes skill and authority.”

While having an owner is essential, it’s not a panacea.  McGovern highlights the skill and time issues associated with ownership.  The problems get even bigger when looked at across the entire organization.  The reality is that content ownership is almost never distributed on the basis of resources available to support the content.  People are content owners because they want political ownership, or are the subject matter expert, or else they were the unlucky inheritor of content no one else wants to own.  Some people are owners of a few items that aren’t burdensome but are largely irrelevant to their main job functions, and hence a low priority.  Other people “own” piles of content they can’t hope to manually review on a regular basis.  Simply devolving responsibility for managing the dead content problem is not a viable solution.

Hope #2: Content reviews

Another tactic that often fails in practice is setting up a content review schedule.  The content owner, who has moral responsibility for keeping his content up-to-date, even if he has no staff hours budgeted to do so, is expected to review all his content on some arbitrary schedule: quarterly, semiannually or annually.  During this review period, all other of their work obligations will presumably pause.  Of course, few modern organizations have fallow periods available for spring cleaning.  But even if an organization did prioritize content review, doing it on an arbitrary timeframe makes no sense.  Such arbitrary schedules are not based on the underlying characteristics of the content, and do not reflect any solid business requirements.

Content varies in its shelf life.  Some content has a short shelf life: it needs to get updated more frequently than once a quarter, and probably does get updated, because it is front of mind.  Other content has a long shelf life.  It may need updating at any time, but when that will be is unpredictable.  If change is infrequent, then reviews will either result in wasted time checking ok content, or allowing a long time to lapse before something that has become outdated is revised.

Content management systems promise to manage the problem by including an expiration date on the content that sends a alert to the content owner.  While tying an expiration date to an item of content does have the advantage of distributing when different content is reviewed, it doesn’t change the the fundamental problem of date-based reviews: they are predicated on an arbitrary timeframe, a guess of when to look at something again.  In the case of a CMS expiration, there may even be a tendency to overestimate how long a piece of content will stay current, since we all like to believe in the lasting value of our efforts.

Scheduled reviews aren’t efficient or effective.  There needs to be better methods tied to actual business requirements.  Use scheduled reviews only as a fall back until you’ve implemented more robust solutions.

Hope #3: Kill unused content

A third tactic, borrowed from the realm of content auditing, is to get rid of content once it’s not being used.  This is essentially a Darwinian approach: if content isn’t being viewed, it isn’t of value to audiences or the organization, so it doesn’t deserve a home any longer.  And, unsurprisingly, much content that isn’t viewed is in fact out of date.  So utilization can be a proxy for out-of-datedness.

Paul Boag suggests this approach as being both labor efficient and skirts some of political squabbles  surrounding content retirement: “An alternative to time based review points would be traffic based. This is designed to remove content that is not really used by users rather than out-of-date content. This review point would be triggered if the traffic to a page falls below a certain threshold over a given period. This would indicate that the page is of little interest and is simply making it more difficult for the majority of people to find what they are after.”

Boag recognizes the human resource bottleneck of manual reviews, and his proposal is elegant.  But killing unused content is both reactive and unstrategic, and could result in some bad outcomes.  The goal should be to keep content up-to-date, not simply to get rid of old or underused content.  Those two goals are not equivalent.

Unfortunately, web traffic is an imprecise measure of the topical value of content.  It merely records what content has been viewed, not the actual interest of audiences in the topic itself.  One can imagine several scenarios where content that should be of interest to audiences and is strategically important, is not getting expected traffic:

  1. Content on important topics may be poorly written; such content should be improved, not eliminated.
  2. The lead subject matter expert for a hot, fast moving topic may have left the organization and could not be replaced right away: the content got old and stale, and traffic fell.  Again, eliminating such content will only weaken public perception of the organization’s expertise about the topic.
  3. There was a SEO problem for the content due to JavaScript technicalities on those pages.  The webpages need fixing here, not the content.

Perhaps more unsettlingly, web traffic analysis may not catch out-of-date content when traffic is high.  Content may be out of date and contain errors, but still gets lots of traffic due to its keywords or placement.  Instead of flagging the problem, the analytics team is blissfully unaware that they have a large scale issue until customer service complaints start coming in.

Analytics should be part of the tool set, but remember that they are looking backward, and can miss some blind spots.  Analytics don’t answer what content the organization should be emphasizing in the future from a business perspective, it only addresses what content audiences have found popular in the past.

Why the status quo needs disruption

The approaches I’ve critiqued do have merits: all are conceptually simple, all are necessary, and in many cases they are the only options organizations currently have to address content that’s gone out-of-date.  Organizations are not wrong for using these approaches.  But relying on these approaches alone will never solve the problem that content goes out-of-date, and organizations only discover this fact after it has happened, sometimes a long time after.  These approaches are about how to resuscitate content that has died, about applying CPR to content that’s stopped breathing.

Every year organizations publish more and more new content, and every year they have more and more legacy content they need to look at to make sure its not out-of-date and in conflict with all the new stuff.  Every year, customer toleration of old content lessens, as people expect faster updates, and get more frustrated encountering legacy content that hasn’t been updated.  A process based on discipline and threats can’t cope with these trends.

To review some of the problems with the status quo:

  • content owners are often not aware their content is out-of-date for long periods
  • other members of the organization may not know of the problem either
  • web analytics won’t necessarily flag the problem

What’s been missing in the content strategy toolkit are ways to anticipate when content will go out-of-date, and make revisions to the content before the content becomes overtaken by new events.  In the second part of this article, I will explore ideas for improving an organization’s anticipatory capability to make content changes when they need to happen.

— Michael Andrews


Making your content relevant

Talk to people who use content, and you will hear complaints about it being too hard to find what they want, about having to deal with extraneous content, and about corporate content sounding like “marketing speak.”  Such complaints reflect a sense that the content people encounter is not relevant to them.  Too much of the wrong content, too little of the right content.

Making your content relevant involves matching what you offer audiences with what audiences want from you.  That sounds simple, but there are several dimensions involved.  Let’s consider how content can match, or mismatch, audience needs.

We will start with a few assumptions to keep the discussion manageable.  We will assume that your brand serves one or more audience segments, but for now we will focus on your relationship with a single segment.  We will assume that the audience segment, while comprised of individuals having multiple interests, can be treated as homogeneous in having a similar set of needs and expectations relating to the content they seek from you.

There are many ways audiences and brands can overlap, both imperfectly and productively.  This can be illustrated in the following diagram.


overlap of audience interests with content offered
overlap of audience interests with content offered

Looking at overlap

The worst outcome is when brands are largely out of touch with their audience.  Audiences feel estranged when they are interested in many things a brand might be able to talk about, but the brand chooses to talk about other things that have little interest to the audience.  The brand could be a trusted source of information on a topic, but instead talks at the audience and offers little content of actual value.  The situation is common, especially in the area of professional services: a bank, rather than provide genuinely helpful suggestions in the context of their products and services, instead promotes messages from the chairman and newsletters about their community involvement.

Another common situation is when audiences use a core of the content you offer them, but ignore the rest.  Audiences are serviced but not engrossed when they value a brand for narrow utilitarian content only.  People will say: “I just want to get an answer to my question; I don’t want to see all this other stuff.”  They engage with the product support content, but not with the content about the brand, because the brand content hasn’t been created in a way that matches the underlying needs of the audience.  People go to their local government website for a schedule of recycling, but ignore articles about public hearings related to the recycling program.  If the audience is not thinking about other things they might potentially want from your brand, you need to rethink your approach to building audience engagement to make your brand feel more inclusive.

People sometimes want more content from a brand than the brand is prepared to offer.  Audiences feel underserved when they need content from a trusted source, but the brand only wants to provide minimal factual content to support sales.  Ordinary people are not experts on tires, water heaters, or many other aspects of life, and want to understand how their situation or behavior will influence product performance in the long term.  But brands may offer only limited specific information, rather than relationship-building content.

Audiences feel supported when there is lots of content available that is relevant to them.  The brand has done a good job translating what it knows into content that speaks the language of the audience and addresses their needs.  From the audience perspective, most of the content seems relevant; from the brand’s perspective, little of the content created is wasted.  Achieving this happy state involves creating compelling audience centric content, where the brand qualities play a supporting role to the content itself.  By having content responsive to audience needs, the brand can build its audience, and have permission to talk about itself where appropriate.

A final possibility is when audiences become obsessive about your brand, and want your brand to take the lead in the relationship.  For brands it is a wonderful position to be in, to have adoring fans who want you to talk about yourself, rather than talk about the humble needs of the people who are interacting with your content.   For these brands the chief challenge is regularly creating highly original content that will impress your fans.  While this fandom brand situation clearly does exist, it is comparatively rare.  Fan-based brands require long term development, and tend to arise in exceptional situations, where extraordinary brand differentiation and identity has been achieved.  While brands should be cautious about directly pursuing this route, it may be possible to incorporate elements of this approach with certain lead customers provided the primarily audience already feels fully supported.

Reshaping your content

To broaden the overlap between audience interests and brand content, shuffle things around.  Enlarge the range of the relevant content:

  1. identify and understand the full range of audience interests through audience research and analytics
  2. build audience curiosity for content that addresses other topics of potential mutual interest by holding content exploration sessions followed up by content testing
  3. create quality content that address these audience interests

Generally this process involves moving toward the audience, away from self-referential content.  Ideally, relevant content is about “us,” but to get there, one needs to address the “me” in the audience before one can talk about “you” the brand.