Self-Service Has a Content Problem, Not a Technology Problem

Only 14% of customer service and support issues are fully resolved in self-service. For issues that customers themselves describe as very simple, the figure is 36%.

Those numbers come from a Gartner survey of 5,728 customers conducted in December 2023. The reflex when leaders see them is to ask whether customers will use self-service at all. The same survey answers that: 73% of customers use self-service at some point in their journey. Usage is not the problem. Completion is.

Where it actually breaks

The survey names the most common failure directly. In 43% of cases, customers could not find content relevant to their issue. A second finding sharpens it: 45% of customers who started in self-service said the company did not understand what they were trying to do.

Read those together and the diagnosis is uncomfortable, because it is not one anybody can buy their way out of. The deflection worked. The customer arrived, engaged and searched. What they found did not match the problem they had.

That is a content and taxonomy failure rather than a platform failure. In practice it usually means the help centre is organised around the organisation’s internal structure, written in the organisation’s vocabulary, and scoped to the questions the organisation anticipated rather than the ones customers actually arrive with. Replacing the search widget fixes none of that.

It is worth being precise about what this means for investment. An organisation in this position does not need a better self-service platform. It needs someone to read the top 200 search queries that returned nothing useful and write the missing answers, in the customer’s words, including the answers that are commercially awkward.

Customers found somewhere else to self-serve

While first-party self-service has been stalling at the content layer, the behaviour moved. A Gartner survey of 5,801 customers conducted in January and February 2025 found that 51% of all customer service journeys now begin on third-party platforms such as Google, YouTube and ChatGPT, with search engines the most popular starting point. Among Gen Z customers, 74% initiate their service journey with a third party before ever reaching an official company channel.

Two further numbers from that survey deserve to be read next to each other. Customers reported an average success rate of 62% in finding what they need on platforms like YouTube or Google. And only 22% of customers start, stay and resolve their issues entirely using first-party company channels.

Keith McIntosh, Senior Principal, Research in Gartner’s Customer Service and Support practice, described third-party platforms as having become the new front door for customer service.

Set the 62% against the 14%. The comparison is not exact, because one measures finding what you need and the other measures full resolution, and I am not going to pretend they are the same metric. But the direction is hard to miss: on their own account, customers are having more luck getting answers about your company somewhere you do not control than inside the channel you built and funded for exactly that purpose.

This is the same dynamic that shows up in third-party AI outpacing company chatbots, and it has the same consequence described in the shift from rankings to credibility. When the front door is somebody else’s, what gets said about you there is assembled from sources you did not write.

What failure actually costs

The usual business case for self-service is cost to serve. The research suggests the real exposure is retention.

A Gartner survey of 6,138 customers conducted in December 2022 found that 38% of Gen Z and millennial customers say they are likely to give up on resolving a customer service issue if it cannot be resolved in self-service. Among baby boomers the figure is 11%. Older customers persist: 55% of them said they would give up only after failing to find an answer having contacted multiple people, against 31% of Gen Z.

Giving up is not a neutral outcome. In the same research, customers reported that they would use the service or product less (55%), would not buy from that company again in future (52%), and would say negative things about the company or product (44%).

So a failed self-service session is not a deflection that did not land. For a younger customer it is a plausible exit point, and the downstream effects land in revenue and reputation rather than in the service budget where the saving was booked.

The same body of research also found that 88% of customer service journeys starting in self-service touch multiple channels, including phone, live chat and email. The handover is not an edge case; it is the normal path, which makes it worth engineering rather than tolerating. Gartner reported that customers who experienced a seamless transition say they are 74% more likely to start in self-service next time they have an issue. That is a relative lift rather than a share of customers, and it is worth quoting carefully, because the effect it describes is a compounding one: a good handover today buys self-service adoption tomorrow.

What to do about it

Audit failure, not traffic. Self-service dashboards report sessions, page views and deflection. None of those distinguish a customer who resolved their issue from one who gave up. The queries that returned nothing, the sessions that ended in a call, and the articles people open and immediately leave are where the diagnosis lives.

Write for the question, not the product. If 43% of failures are missing or irrelevant content, content is the project. That means structuring around customer intents and phrasing rather than product lines and internal ownership, which is journey work applied to a knowledge base.

Treat the third-party front door as part of your estate. If half of journeys start on Google, YouTube or an AI assistant, then what those surfaces say about your product is part of your service experience whether or not anyone in the service function owns it. Most organisations have nobody accountable for it at all.

Engineer the handover. Since 88% of self-service journeys touch another channel, the transition is the main event. Carrying context across so the customer does not restart is the single highest-leverage fix available, and it is usually blocked by system boundaries rather than by cost.

Measure resolution and effort, not containment. Containment rewards ending the session. Resolution and effort measures reward finishing the job, and they are the only metrics that would have caught a 14% resolution rate before a customer survey did.

What this data does not tell you

These are three separate Gartner surveys with different samples and fieldwork dates: 5,728 customers in December 2023, 5,801 in January and February 2025, and 6,138 in December 2022. They are not a time series and should not be read as one. None publishes a GCC breakdown, so nothing here is regional evidence; anyone applying it to this market, including me, is extrapolating. The 2023 Q&A release is also published in a question-and-answer format rather than as a standard announcement.

I also deliberately left out a widely circulated figure claiming 74% of customer service journeys involve multiple channels. It appears only in gated Gartner research pages with no published sample size or fieldwork dates, and it gets confused with the unrelated 74% figure about seamless transitions. The 88% statistic above is narrower, is specific to journeys starting in self-service, and is published openly.

The summary

Self-service is not failing because customers refuse it or because the technology is immature. Nearly three quarters of customers use it. It fails because, at the moment of need, the answer to their particular question is not there in language they recognise.

That is a less exciting problem than the one most roadmaps are built around, and considerably cheaper to fix. It is also the one the evidence actually points at.

Martin Alva

Marketing, Brand & Digital Transformation Leader, United Arab Emirates

Senior Manager, Employee Experience (Brand Positioning, Digital Platforms & AI) at Space42 (A G42 & Mubadala Company). 15+ years in marketing, brand and digital transformation across MENA and Europe. 5 MENA Effie Awards.

More about Martin Alva → Connect on LinkedIn →