The rating is bad, the complaints repeat, and the founder reaches the standard conclusion: we need someone on support. Sometimes that is right. More often it is the most expensive way to buy the wrong fix, because a hire adds capacity to whatever process exists, including a broken one. Faster answers into the same dead ends produce politer versions of the same reviews. Before the job listing goes up, here is how to tell which problem you actually have.
The diagnostic: capacity problem or system problem?
Pull your last thirty negative reviews and your support queue, and sort what you see into two piles:
Signs you have a capacity problem (the hire helps)
- Response times are the complaint: "took five days to hear back" appears verbatim in reviews.
- Resolutions are good when they finally happen: customers who get through leave satisfied.
- Ticket volume per agent is objectively unreasonable, and peaks bury the queue for weeks, the pattern from the Black Friday piece.
Signs you have a system problem (the hire hides it)
- The same failure repeats: sizing complaints, tracking black holes, refund delays. Recurring themes are process defects wearing customer faces; answering them faster does not stop their production, the point we made in the 21-point audit.
- Agents cannot actually fix things: every reship or refund needs founder approval, so speed dies at the authority bottleneck regardless of headcount.
- The public record is untouched: nobody owns review replies, nobody runs collection. In our study of 578 damaged brands, 80% collected nothing and 56% answered nothing, and those are not staffing numbers, they are ownership numbers: tasks with no owner do not get done at any headcount.
- Happy customers are silent: the rating problem is an asking problem, and hires answer inboxes, they do not redesign post-purchase flows.
What moves a rating, ranked by cost-effectiveness
For a store whose rating is the presenting symptom, the interventions rank consistently:
- 1. Install collection on every order. Delivery-anchored, honest, permanent, per the timing guide. Cost: configuration. Effect: the silent satisfied majority finally counterweights the vocal angry minority. Nothing a hire does compares.
- 2. Answer the public backlog. Every negative, this week. Cost: a hard afternoon with the templates. Effect: the profile reads as inhabited to every future checker, human and machine.
- 3. Give the frontline authority. Decide what support can grant without asking (reship, label, partial refund, threshold). Cost: one founder decision. Effect: resolution speed doubles at existing headcount.
- 4. Fix the top recurring defect. Whatever theme leads your negative pile: the size chart, the courier, the refund lag. Cost: varies. Effect: turns off a complaint factory at the source.
- 5. Then, add people to a working system. When items one to four exist and the queue still overflows, the hire multiplies something worth multiplying, and the job description writes itself from the system's actual gaps.
The uncomfortable middle option
Between "hire someone" and "do nothing" sits the option founders avoid naming: run the system yourself for one quarter. Not forever, and not because your time is free, but because the founder is the only person who can make items three and four happen, and the recovery conversations, the ones that turn angry customers into the best reviews per the recovery piece, carry a weight from the founder that no agent matches. A quarter of founder-run reputation work typically fixes the system and documents it, at which point delegation, internal or external, inherits something that works.
And that is the honest framing for what we sell too: not staffing, but the system installed and run, with the store's own voice, until the rating reflects the actual business. Whether your situation is a capacity gap, a system gap, or both, is exactly what the free audit is built to tell you, before you pay a salary to find out.
FAQ
Our support inbox is drowning right now. Surely the hire comes first?
Triage first: authority (item three) and the top-defect fix (item four) shrink queues within weeks by removing repeat tickets. Hire into the queue that remains, not the one you have today.
Can a virtual assistant run review replies and collection?
Execution yes, ownership no: the tone rules, the authority thresholds and the escalation lines need defining first, or you get the gratitude-robot profile described in the positive replies piece.
What does a reputation-focused role cost versus an agency?
A capable dedicated hire costs a multiple of any agency retainer, which is the wrong comparison anyway: the question is whether a working system exists for either to run. Systems first, then the cheapest competent operator of them.
How fast should the rating move once the system is in?
Profile readability changes in weeks, the number follows order volume over months, as detailed in the frozen reputation piece. Anyone promising faster is describing methods this blog exists to warn you about.