A customer calls, emails, or chats, and the question was always: is there someone ready to pick up. Reachability was thereby a function of staffing. More capacity on the phone or in the inbox meant a better score on this dimension. Whoever was staffed in the evenings and on weekends beat whoever was not. That mechanism was fixed and could be copied by competitors simply by scheduling more people.
That mechanism is no longer the only thing that counts. When part of the first contact is handled by AI, reachability shifts from a staffing question to a question about which steps in the contact process have been automated, which steps run with oversight, and which steps remain human work. Whoever has organized that distinction sharply wins on this dimension in a different way than whoever mainly hires extra people.
Reachability does not consist of one task but of a chain of steps, and those steps do not all shift at the same pace.
What AI can take over completely. The initial recognition of a question, returning standard information, confirming an appointment, routing to the right department: these are tasks with a fixed pattern and a limited number of outcomes. At a growing number of companies this already runs outside office hours without an employee monitoring it. At other companies the same process still rests entirely on people, not because it couldn't be done, but because it has not yet been set up.
What happens partly with human oversight. As soon as a question deviates from the standard pattern, touches a complaint, or a commitment has financial consequences, the work shifts to an intermediate form. AI proposes an answer or a classification, and an employee approves or rejects it, with reason. This is the point where companies differ most from each other: how much of the traffic passes through this layer, and how quickly the approval proceeds, determines how quickly a customer is ultimately helped.
What remains human work. Complex negotiation, escalations where relationship and context weigh more heavily than the protocol, and situations with legal or financial weight remain with people. That is not a temporary limitation of the technology but a choice about where risk and customer relationship weigh more heavily than speed.
The division between these three layers is not fixed. It shifts per company, per type of question, and per moment, and it is precisely this shift on which competitors are now diverging.
A customer does not notice whether a response was drafted by AI or by a person. A customer notices whether an answer comes, how fast, and whether the answer is correct. Response time outside office hours, the first-time-right score of an answer, and the extent to which a question reaches the right layer without waiting time: those are the signals that make reachability distinctive today. A company that has automated recognition and standard handling and has properly set up the oversight of it can respond faster and more consistently than a company that deploys just as many people but processes everything manually.
Whether this shift has already taken place depends on the type of customer question, on how repeatable the process is, and on how much volume runs through it. A company with many similar questions has a stronger case for automating the first layer sooner than a company with mostly unique, context-rich conversations. That is also why two competitors in the same market can be far apart on this dimension without one having more people than the other.
This shift concerns how work is organized, not who is allowed to keep doing that work. Decisions about personnel are subject to their own legal requirements; what counts here is which steps have demonstrably shifted and what that means for the score on reachability.
Reachability does not stand alone. A company that is ahead on digital maturity usually also has the systems in place to automate the recognition layer. Reach in the sense of where a company can serve customers touches on geographic coverage, because being reachable 24/7 partly compensates for the absence of local presence. And whether a claim about fast response time holds up against a specific competitor can be tested with a peer group composed on the right points of comparison.
The underlying question of which work in this company can truly be taken over by AI is answered per task with the work scan from FTE TO AI.
Reachability is a dimension on which claims are easily made and difficult to prove: everyone says they respond quickly, few companies can substantiate that with figures per layer. Do you want to know whether your lead on reachability holds up, or whether a competitor is by now responding faster through a different division between automation and oversight: name where you think you are winning and do the free dimension check. You will then see which of those claims can be defended with evidence. For companies that want to look more broadly than their own assumption, there is also an explanation of the white space analysis, to see which points of comparison remain unused. The full benchmark, with the evidence matrix across all dimensions, is under construction.