Customer Complaint and Redo Visit Standard

Purpose

Make a redo a measured event rather than a free re-clean. A shop that sends someone back whenever a customer is unhappy keeps the customer and learns nothing: by quarter end it has a redo rate, a tired crew and no idea whether it is training badly, selling the wrong scope, or scheduling too little time. What separates those three is a written difference between what was flagged and what the visit was sold to include, captured at the complaint and again at the redo.

Recovery is also mostly response speed, and a shop with no stated window improvises one per customer, which is how a good account learns it has to escalate.

Scope

Covers a quality complaint on completed work: missed areas, surfaces the customer reads as not clean, streaking, odor, or a finish they expected and did not get, from intake through the decision, the redo visit and the record.

Excludes: which flagged area to redo first (the redo triage tree); the cause of a specific defect (the streak, haze and residue trees); damage the crew caused (the breakage SOP); a missing item (the found-valuables SOP); and ending an account (the complaint resolution tree).

Roles and responsibilities

Procedure

1. Capture the complaint in the customer's own words and start the clock at first contact. Write what they said, not your interpretation, plus the room, the surface and when they noticed. Acceptance: a verbatim line per flagged item, the visit date, and a timestamp on the intake. Wrong looks like "customer unhappy with bathrooms", which cannot be compared to anything later. Stop rule: an intake that cannot name a room and a surface goes back for one clarifying question before anything is scheduled. Hazard: none, a phone call.

2. Acknowledge inside a stated window, before you know the answer. A workable default is acknowledgement within one business hour of the complaint arriving, with the decision by end of the next business day; tune to your own staffing, but publish the numbers. Acceptance: the customer has heard from a named person inside the window, with a specific time for the decision. Wrong looks like silence while the office investigates, which the customer reads as being ignored. Stop rule: if the window will be missed, call anyway and say when; a missed window that is announced costs a fraction of one that is discovered. Hazard: none physical; the risk is a first responder conceding fault before the scope is pulled, so step 3 comes before any promise.

3. Compare the flagged items against the sold scope before you promise a redo. Pull the service agreement or the account's task list and mark each flagged item in scope, out of scope, or ambiguous. Acceptance: every flagged item carries one of those three marks, in writing, before the customer is called back. Wrong looks like a manager agreeing to a redo on an item never sold, which teaches the customer the scope is negotiable and guarantees the same call next month. Stop rule: an ambiguous item is treated as in scope for this visit and the scope document is fixed the same week; a clearly out-of-scope item is not a redo, it is a quote, told plainly and offered. Hazard: none, desk work.

4. Set the complaint window on the account and apply it consistently. A workable default: complaints accepted within twenty-four hours of a recurring or standard visit and within seventy-two hours of a deep, move-out or post-construction clean, which get inspected later by a landlord, agent or buyer. Acceptance: the complaint's age against the window is written on the intake, and the same numbers apply to every account. Wrong looks like a window enforced on small accounts and waived for large ones. Stop rule: a complaint outside the window is still worked, but as a goodwill visit coded separately, so it does not distort the redo rate in step 8. Hazard: none; the failure mode is a metric that stops meaning anything.

5. Choose who goes back on evidence, not on blame. Default is the original cleaner with the lead present, because the person who missed it learns most by seeing it; the exceptions are a customer who asked for someone else, a second complaint against the same cleaner on that account, and any complaint about conduct rather than quality. Acceptance: the decision and its reason are one line on the ticket. Wrong looks like the newest cleaner sent to fix someone else's work, which trains nobody and reads to the customer as a downgrade. Stop rule: a conduct complaint never routes back to the same person; it goes to the owner. Hazard: none at the desk; on site, a cleaner who feels accused rushes, and a rushed redo is where step 7's chemistry mistakes happen.

6. Photograph the flagged surfaces before touching them, then work the ordered list. One frame per flagged item before, one after, from the same position and in the same light. Acceptance: a matched before and after pair for every in-scope item, plus the customer's confirmation on site if present. Wrong looks like an after set only, which proves the surface is clean now and nothing about what was found. Stop rule: if a flagged surface is already clean on arrival, photograph it, say so to the customer, and do not re-clean it to be agreeable; a redo on a surface that passed teaches the shop the wrong lesson from its own record. Hazard: kneeling and reaching into the same tight bathroom or oven spaces twice in a week is where the strain and the cut happen, so the redo gets the same PPE and tools as the first pass, not a bare hand and a corner of a cloth.

7. Do not escalate the chemistry to make the redo look decisive. Work the same product family as the original visit, at its labeled dilution, unless the service manager has approved a change and the surface can take it. Acceptance: the product and dilution are written on the ticket and either match the original visit or carry the approval. Wrong looks like a cleaner reaching for a stronger acid or a bleach product on a surface that carried a quat or an alkaline cleaner days ago. Stop rule: any product change on a redo stops for the service manager, who reads section 10 of the new product's safety data sheet against what the first visit left on that surface before approving it, and that sheet has to be accessible to the crew on shift because 29 CFR 1910.1200 requires it for every hazardous product the shop puts in their hands; any residue of unknown identity on the surface stops the work entirely until the customer says what they used. Hazard: the real one on a redo. Mixing an acid product over a chlorine bleach residue liberates chlorine gas, and mixing bleach with an ammoniated cleaner liberates chloramine; both are inhalation hazards in a small unventilated bathroom. Both warn you first, and the warning is the instruction: chlorine is detectable by smell well below the level that injures, and chloramine stings eyes and airway on the first breath, so a sharp or irritating smell that was not there a moment ago means leave the room now, door shut, and do not go back in to open a window or cap the bottle. Get outside into fresh air, and a cough or chest tightness afterward is seen by a clinician that day rather than waited out, because injury from these gases can arrive hours later. Prevention is the rest: ventilate, one chemistry at a time, and rinse between.

8. Close the loop with a root-cause code and a change that is not a conversation. Code every redo to a short fixed list: missed area, technique, time under-allocated, scope mismatch, customer-caused between visits, or product or equipment fault. Acceptance: one code per complaint, the customer told what changed, and the change written on the account or the route: minutes added, a task moved into scope, or a tool replaced. Wrong looks like a redo closed with "spoke to the cleaner", which is not a change and will not survive turnover. Stop rule: the same code appearing three times on one account, or three times for one cleaner in a month, stops being a coaching item and goes to the owner as a system item. Hazard: none; the failure mode is a shop that runs redos indefinitely believing it is fixing them.

The record this produces

The complaint record carries the verbatim intake with a timestamp, the scope mark per flagged item, the window age, who went back and why, the matched photograph pairs, the product and dilution used, the root-cause code, and the change made.

The service manager reads the file to close the visit; the owner reads it monthly on two cuts. Redo rate is redos divided by completed visits that month; pick a line and hold it, and a workable start is three percent, tuned to your mix, since a move-out heavy month runs higher. Code distribution is the more useful cut, because it names the kind of problem: scope mismatch is a sales and paperwork fix, time under-allocated is a route fix, technique is training, and a cluster on one account is usually neither.

Worked pass

Recurring biweekly, standard 105-minute visit. The customer emailed the morning after service listing three items: the guest bathroom mirror streaked, the kitchen baseboards dusty, and the oven interior not cleaned.

Step 1 captured all three verbatim at 8:40. Step 2 acknowledged at 9:15, thirty-five minutes in, inside the one-hour window, decision promised by end of the next business day.

Step 3 failed on the third item. The comparison marked the mirror and the baseboards in scope, and the oven interior out of scope, because it is a separately quoted add on this account and never included. The stop rule ran on that item as written: the oven was not scheduled as a redo. It was named on the callback as a task the visit does not include and quoted as an add for the next visit, which the customer took.

Step 4 put the complaint about seventeen hours after service, inside the twenty-four-hour recurring window, so it coded as a redo rather than goodwill. Step 5 sent the original cleaner with the lead, first complaint on this account, reason written.

Step 6 produced two matched pairs: the mirror with drying streaks in window light, then no visible streak at the same angle; the baseboard with dust along the top edge, clean after. Step 7 used the same glass product and neutral cleaner at their labeled dilutions, no change, both written on the ticket; the bathroom fan ran throughout. The redo ran thirty-five minutes, one third of the 105-minute original visit.

Step 8 coded it time under-allocated rather than technique: both misses were end-of-visit tasks and the route had this house at the end of a four-stop day. The change was fifteen minutes added to the account's standing allocation. That month the shop closed four redos against 118 completed visits, 3.4 percent, above the three percent line, so the owner ran the code distribution: three of the four coded time under-allocated, all three last stops. The fix was to the route, not to three cleaners.

References

• Product labels for every chemistry used on a redo. Contact time and any incompatibility statement are on the label. For a registered antimicrobial, meaning a disinfectant or sanitizer, use inconsistent with the labeling violates FIFRA; for a non-registered cleaner such as a glass or neutral cleaner, the label is the manufacturer's direction rather than a legal instrument, though it still governs dilution and surface.
• 29 CFR 1910.1200, hazard communication, which puts the safety data sheet step 7 checks in the crew's reach on shift.
• See related: cleaning-several-flagged-areas-which-redo-first-triage-decision-tree for ordering the flagged list; cleaning-customer-complaint-resolution for when an account should end; cleaning-walkthrough-and-final-inspection for the inspection that prevents most of these; cleaning-chemical-mixing-ppe-protocol for what may never be combined.