How to Retain Commercial Cleaning Clients
An inspection scorecard, a complaint routine and a renewal calendar, with the evidence labelled by strength.

In this article
- Why accounts leave, and what the evidence can and cannot tell you
- Agree what "good" means before you inspect
- The inspection loop
- The scorecard
- One contact and a communication rhythm
- Complaint resolution and service recovery
- Early-warning signs
- Account reviews and contract renewals
- What retention is worth
- When an account leaves anyway
- A first 90 days of a retention routine
- Sources
Picture a Monday when a property manager you have served for two years asks, in passing, for a copy of your contract "for the files." You send it. Nothing else changes, and the invoices keep getting paid. A trade-association writer lists that exact request among nine early signs that an account may be going out to bid, with the caveat that it can be routine and that no single sign proves anything. What matters is the position this leaves you in. If you learn about a problem from a request for proposals, most of the time you needed is already gone.
Keeping commercial cleaning clients is mostly a matter of knowing, early and in writing, how the work looks to the person who pays for it. This article lays out a way to do that: a scored inspection, one named contact and a communication rhythm, a routine for complaints, a short list of warning signs, and a renewal calendar that begins months before the contract ends. It includes an inspection scorecard you can copy, with a worked sample.
It is general education for owners with roughly three to thirty commercial accounts. It describes practice, labels where the evidence is thin, and promises no result. Nobody can guarantee that a client stays.
Why accounts leave, and what the evidence can and cannot tell you
A common assumption is that clients leave over price or over a single bad night. The evidence we could verify does not settle that, and it is worth being plain about how little there is.
The best number we found comes from the 2024 Building Service Contractor Market Study, published by Contracting Profits and BSCAI (the Building Service Contractors Association International). It was an email survey of 14,643 contractors drawn from magazine subscribers and BSCAI's database, answered between January 30 and February 18, 2024. There were 169 respondents, with a stated margin of error of 7.54 percent at the 95 percent confidence level. Asked what account turnover they experienced each year on average, 73 percent chose under 10 percent, 12 percent chose 11 to 15 percent and 6 percent chose 16 to 20 percent. The report's introduction summarizes this as 91 percent of contractors seeing less than 20 percent client turnover.
Read that with its limits. Respondents chose to answer, came from trade-publication and association lists, and served an average of 191 accounts, which describes an established firm, not a company with a dozen buildings. One survey, one year. We would not call it the industry retention rate, and the study does not claim to be one. It is a reference point for how established contractors described their own turnover.
The same survey asked contractors what is most important to their customers. The answers were a clean appearance (44 percent), a healthy and sanitary environment (37 percent), a low price (18 percent) and a green cleaning program (1 percent). These are contractors' views of their clients, not the clients' own answers, and they say nothing about why accounts are lost. They do suggest that visible results carry more weight with the people contractors serve than price alone does, which is a reason to measure what the client can see.
Vendor articles repeat larger claims about retention, such as multipliers on acquisition cost or profit. We could not trace them to a source that studied cleaning, so we leave them out. The reasoning needs no multiplier: a recurring account that earns a margin each month is worth keeping, and replacing it takes the sales effort described in our article on why cleaning companies shouldn't rely on ads alone.
Agree what "good" means before you inspect
An inspection measures service against a standard, and the standard lives in the scope of work: which areas, which tasks, how often. Our guide to starting a cleaning business and the piece on contracts and service agreements cover how to write one. If the scope says only "office cleaning," an inspector and a client can look at the same desk and disagree.
ISSA, the cleaning industry association, publishes the Cleaning Industry Management Standard (CIMS). Its 2023 edition says an organization "shall" measure and document performance against the scope of work, using one or more of surveys, inspections, complaints or customer evaluations. By the standard's own definitions, "shall" is mandatory for full compliance, "should" is a recommended element of which only a percentage need be met, and "may" is advised or suggested but not a requirement. CIMS is an industry standard, not a law, so a company with no plan to certify has no obligation to follow it. It is a useful outline of what a quality system contains, and we keep its verbs intact below.
CleanLink, a trade publication, ran a 2015 column by consultant Mickey Crowe suggesting a pass mark for each area, set from the contract and stricter for high-visibility spaces. His example is an executive office that may need 100 percent while a back work area may only need to be satisfactory at 85 percent. He also advises listing expectations that go beyond the contract, then either revising the specifications or explaining to occupants what the service covers. A complaint about something the contract never included is a pricing conversation, and treating it as a performance failure wears down your crew and your margin.
The inspection loop
CIMS describes inspection in layers. Operational inspections by cleaning staff "should" be done as service is delivered, site supervisors "should" conduct site inspections, management "should" conduct unannounced site inspections, and customers "may" take part in management-level ones. The standard gives no frequency, and no source we read does either, so the schedule below is the Journal's opinion.
A workable rhythm for a small company: the cleaner checks their own work against the site checklist before leaving. The owner or supervisor inspects every restroom and a rotating sample of the other areas, weekly or every other week for the first 90 days at a new site, then monthly. Once a quarter, the client's contact joins a walk-through.
Three habits help. Inspect after the crew finishes and at varied times, since a site checked only on a predictable schedule shows how the crew performs when watched. Score areas and sites, not named individuals: scores attached to a person can feed discipline and pay decisions, and employment rules differ by state, so ask an employment attorney or your state labor agency before the numbers reach a personnel file. And photograph only with the client's agreement, because access agreements and privacy rules, health-care ones especially, can restrict it.
The scorecard
What follows is the Journal's own framework. It is not an ISSA, BSCAI or APPA standard, and no trade body endorses it. The categories, weights and score bands are judgment calls made to be adjusted. Change the weights to match each contract's written scope, and keep the categories fixed for a given site so that scores can be compared over time.
The scale. Score every line item from 0 to 4.
| Score | Meaning |
|---|---|
| 4 | Meets the agreed standard |
| 3 | Minor deficiency: small, and easily fixed during the visit |
| 2 | Moderate deficiency: visible to a client, needs a fix before the next service |
| 1 | Major deficiency: clearly not done, or done badly |
| 0 | Not done |
| N/A | Not in scope at this site; leave it out of the calculation |
The form. Copy this table into a spreadsheet or print it. Delete line items the contract excludes and add any it includes. Each category should have three to five items.
| Category (default weight) | Line items to score 0 to 4 |
|---|---|
| 1. Restrooms (25) | Sinks, mirrors and fixtures; toilets and urinals; floors and drains; dispensers stocked; odor and trash |
| 2. Floors (20) | Hard-floor appearance; edges and corners; carpet vacuuming; entrance mats |
| 3. Entrance, lobby and common areas (10) | Glass and doors; reception and furniture surfaces; trash and general appearance |
| 4. Offices and meeting rooms (15) | Desk and table surfaces, as scoped; trash emptied; dusting, high and low, as scoped; chairs and furniture |
| 5. Kitchens and break areas (10) | Sinks and counters; appliance exteriors; floors and trash |
| 6. Trash, recycling and consumables (5) | Liners replaced; consumables stocked |
| 7. Scope and schedule compliance (10) | Daily tasks complete; periodic tasks done on schedule; service log complete |
| 8. Safety, security and conduct (5) | Keys and alarm procedure followed; wet-floor signs and chemical storage; janitor closet tidy |
The weights add to 100. Restrooms and floors carry the most because they are where building occupants tend to form an opinion, but that is our judgment: the survey above does not rank rooms.
The arithmetic.
- Category score (percent) = sum of line-item points ÷ (4 × number of items scored) × 100.
- Overall score = the sum of (category weight × category score) ÷ the sum of the weights of the categories scored.
- If a category is N/A at a site, drop its weight from both sums and write down which categories were excluded.
Rating bands (Journal defaults, adjustable by contract).
| Overall score | Reading | Action |
|---|---|---|
| 90 to 100 | Meets standard | Record, tell the client |
| 80 to 89.99 | Acceptable | Issue a fix list |
| 70 to 79.99 | Below standard | Corrective plan, re-inspect within 7 days |
| Below 70 | Failing | Owner calls the client's contact that week |
Two overrides, so an average cannot hide a problem. Any single category below 70 percent triggers its own corrective action even when the overall score passes. Any critical item, such as a safety hazard, a security lapse such as access left open, an unusable restroom or a biohazard not handled, is recorded as a critical failure regardless of score and answered within 24 hours.
| Category | Weight | Item scores | Points | Max | Category % |
|---|---|---|---|---|---|
| Restrooms | 25 | 4, 3, 4, 3, 4 | 18 | 20 | 90.00 |
| Floors | 20 | 3, 2, 4, 4 | 13 | 16 | 81.25 |
| Entrance, lobby, common | 10 | 4, 3, 4 | 11 | 12 | 91.67 |
| Offices and meeting rooms | 15 | 3, 4, 2, 3 | 12 | 16 | 75.00 |
| Kitchen and break area | 10 | 3, 2, 3 | 8 | 12 | 66.67 |
| Trash, recycling, consumables | 5 | 4, 4 | 8 | 8 | 100.00 |
| Scope and schedule | 10 | 4, 3, 3 | 10 | 12 | 83.33 |
| Safety, security, conduct | 5 | 4, 4, 3 | 11 | 12 | 91.67 |
Weighted sum: (25 × 90.00) + (20 × 81.25) + (10 × 91.67) + (15 × 75.00) + (10 × 66.67) + (5 × 100.00) + (10 × 83.33) + (5 × 91.67) = 8,375.00 using unrounded category percents. Divide by the weights, 100, for an overall score of 83.75. That falls in the "acceptable" band, so a fix list goes out. The kitchen, at 66.67, is below 70, so the category override requires its own corrective action even though the site passed overall.
Now suppose the same building has no kitchen and that category is marked N/A. The remaining weight is 100 − 10 = 90, and the weighted sum is 8,375.00 − 666.67 = 7,708.33, giving 7,708.33 ÷ 90 = 85.65. The score rose only because the weakest category left the calculation. That is why the form records exclusions, and why a site's categories should not change between visits without a note.
What to record at every inspection. Date, inspector, areas sampled, every line-item score, photographs where the client permits them, issues found, the date by which each will be fixed, who fixed it, and the result of the re-inspection. Then keep a trend table for each site:
| Inspection date | Overall | Lowest category | Critical items | Fix list closed? |
|---|---|---|---|---|
| (date) | (score) | (name and score) | (count) | (yes or no, date) |
If overall scores of 83.75, 88.0 and 91.5 appeared on three successive inspections, the change is 91.5 − 83.75 = 7.75 points. Those three numbers are invented for illustration.
What the scorecard cannot tell you. It records what an inspector sees at one time. It does not capture what the client's staff experience at 7 a.m., a missed supply or a change of mood on the client side. A site can score 92 and still lose the account, so use the number to start conversations with the client.
One contact and a communication rhythm
BSCAI's "5 Tips for Managing Contract Cleaning Customer Relationships" (March 2026) interviews three contractors. None of its advice is quantified, and it reflects the experience of those three firms. Several points still translate to a smaller firm. One contractor's CRM flags accounts where communication or inspection frequency has slipped, with a goal of a monthly touchpoint for every client. Another uses 30, 60 and 90-day check-ins for new clients and moves to quarterly or semiannual reviews once the account is stable, and shares quality-control findings with clients so that the work is documented, observing that most people do not want to talk about cleaning unless there are service issues. A third separates operations from a client-services function, which gives the client one consistent point of contact. The article also says to keep contacting clients who go quiet, by more than one channel, and not to wait for replies.
For a company at your size, that translates to a rhythm like this. Name one person on your side who owns each account, and tell the client who it is. Check in at 30, 60 and 90 days after a start, then each quarter. Keep a log of every conversation and every issue, available to anyone who touches the account. Where an inspection was done, send the result, including the deficiencies, before the client finds them. Consultant Nicole Bernardo, in a 2025 BSCAI article, describes a contractor's 24-hour response promise for service tickets and a 15-minute monthly client call; that is one author's experience, not a measurement. A quarterly meeting agenda might look like this:
Account: [client and site] Date: [date] Our owner: [name] Client contact: [name] 1. Inspection results since last review: overall score [ ], lowest category [ ], critical items [ ] 2. Issues raised since last review, and what was changed to stop a repeat 3. Changes at the client: people, hours, space use, budget, plans to move or expand 4. Anything the client expects that the contract does not list 5. Contract end date: [date]. Renewal conversation to start by: [date] 6. Agreed actions, owner, date
The fourth item has a second use. Our article on finding the person who hires cleaning companies describes how decisions differ between buildings. Once you hold an account, apply it: learn who else has a say, such as a regional manager, a property owner or a purchasing group. BSCAI's warning-signs article notes that a new decision-maker may rebid vendors, so an account known only through one person is exposed to that person leaving.
Complaint resolution and service recovery
A complaint is an inspection performed by the client, and it arrives without a score sheet. CIMS says an organization "should" keep a record of complaints, keep the complainant's identity confidential as far as possible, and document what changed to reduce the chance of a repeat. It also says the organization "shall" respond to positive and negative feedback and inspection results and "shall" fully investigate complaints, and that management "should" meet the customer's designated representative to review results and complaints, then report back when corrective action is taken and ask for feedback. That is the whole cycle in five verbs: record, acknowledge, investigate, correct, report.
In practice, the steps run as follows. Acknowledge in writing the same day, even if only to say who is looking into it and when you will report back. Find the cause before promising a fix: a missed task points to supervision or training, a consistently late finish may point to a production rate that was priced too optimistically, and a complaint about something outside the scope is a scope conversation. Fix the problem, then tell the client what was done and what changed so it does not recur, and ask whether it resolved the concern. Record the complaint, the cause and the prevention step in the account log.
Subject: Your note about [area] at [site] Hello [name], Thank you for telling us about [issue]. We have it logged as of [time, date], and [name] is looking into it today. We will write to you by [time, date] to say what we found and what we are doing about it. If anything else comes up before then, reach me directly at [phone]. [Your name], [company]
In the March 2026 BSCAI article, one interviewed contractor says speed has delivered the best return for the firm. That is a statement of experience, not a measurement, but it fits the routine above: a same-day reply costs little.
Early-warning signs
Kristina Thayer's October 2026 BSCAI article, "9 Scary Signs Your Account May Be Going Out to Bid," is opinion from a contractor-side author, and it says plainly that none of the signs alone mean a client is going out to bid, while two or three together deserve close attention. In our words, the nine are:
- Communication tapers off, with reviews postponed and no feedback either way.
- A new decision-maker arrives.
- A merger or acquisition puts the client's vendors under review.
- Someone asks for a copy of the contract, especially its term and termination provisions.
- Talk about budget pressure and benchmarking increases.
- Add-on and one-off work stops.
- A refresh of the scope of work is requested without an operational reason.
- Small issues, once solved by phone, start to be documented in writing and enforced to the letter.
- Unfamiliar visitors appear, procurement takes over communication, or someone asks for key logs or access lists.
Treat the list as a reason to look at an account, not a verdict. A practical use is a line in each quarterly review log: which of the nine, if any, did we notice? An account showing two signs gets a visit from the owner, an inspection with the client's contact and an early start on the renewal work below.
Account reviews and contract renewals
Kristy Elmore of Harvard Maintenance wrote in a September 2025 BSCAI article that renewal preparation should begin well ahead of the end date: at least four to five months, or roughly 120 to 180 days, particularly for larger clients. She describes the renewal as a relationship review, scheduled in advance, with questions about candid feedback, changes in key stakeholders, shifts in how the building is used and budget pressure. She recommends gathering inspection scores, response times, complaint rates, labor hours against budget and supply use, and, if possible, connecting the service to the client's own outcomes, and suggests a standard playbook or checklist so everyone stays aligned.
For a contract ending on June 30, 2027 (an illustrative date), counting back gives these markers:
| Days before end | Date | Action |
|---|---|---|
| 180 | January 1, 2027 | Pull inspection scores, complaint log, hours versus budget; check the margin |
| 120 | March 2, 2027 | Hold the relationship review with the client; ask for candid feedback |
| 90 | April 1, 2027 | Send the renewal proposal, including any price change and the reason |
| 60 | May 1, 2027 | Follow up; settle remaining questions |
| 30 | May 31, 2027 | Signed renewal in hand, or a clear plan for what happens next |
The 120 to 180 day window comes from the BSCAI article; the action assigned to each row and the dates after day 120 are our suggested schedule.
The margin check. Before proposing, compare the account's actual costs with the price. If wages have risen, the price should change, with notice, which our flagship guide lists among the steps of relationship management. Here is the arithmetic for one illustrative case with invented numbers. A contract is priced at $10,000 a month. Labor including burden is 60 percent of price, $6,000. Other direct costs are 10 percent, $1,000. Gross margin is $10,000 − $6,000 − $1,000 = $3,000, or 30 percent. Suppose labor costs rise 5 percent and the other costs stay put.
- New labor cost: $6,000 × 1.05 = $6,300, an increase of $300 a month.
- If the price stays at $10,000, margin becomes $10,000 − $6,300 − $1,000 = $2,700, which is 27.0 percent. Over twelve months the unrecovered cost is $300 × 12 = $3,600.
- To hold the 30 percent margin: total cost is $7,300, and $7,300 ÷ 0.70 = $10,428.57, an increase of 4.29 percent. Check: 30 percent of $10,428.57 is $3,128.57, and $7,300 + $3,128.57 equals $10,428.57.
- To hold only the $3,000 dollar margin: price = $7,300 + $3,000 = $10,300, an increase of 3.0 percent, and the margin falls slightly to 29.13 percent.
The comparison shows the choice you are making. Because labor is 60 percent of the price in this example, a 5 percent wage rise calls for a price increase below 5 percent. Which target to hold is a business decision, and our pricing article and the bid calculator show how to rebuild a price from current costs. The calculator turns your inputs into a price per visit and a monthly price at the margin you choose. It cannot tell you whether a client will accept the increase.
Whether you may raise a price during a term, require notice, or renew automatically depends on the contract's wording and on state law. We did not research those rules. Have an attorney review your template, and treat everything here as general information, not legal advice.
What retention is worth
Three measures computed from a month-end client list separate problems that one "retention" number hides. All figures below are an illustration with made-up numbers. Take a company with 20 accounts and $30,000 in monthly recurring revenue at the start of a year. During the year three accounts cancel, taking $4,200 of monthly revenue. Two surviving accounts add a combined $900 of monthly revenue through price increases and extras. Four new accounts add $5,600, which the retention measures exclude.
- Account retention = (20 − 3) ÷ 20 = 85 percent. Account churn is 3 ÷ 20 = 15 percent.
- Gross revenue retention = ($30,000 − $4,200) ÷ $30,000 = 86.0 percent. It counts only what was lost.
- Net revenue retention = ($30,000 − $4,200 + $900) ÷ $30,000 = 89.0 percent. It also credits growth within the surviving accounts.
- The year ends with 20 − 3 + 4 = 21 accounts and $30,000 − $4,200 + $900 + $5,600 = $32,300 in monthly revenue.
The 15 percent churn sits inside the "under 20 percent" range most respondents reported in the 2024 survey, but the survey's limits apply in full, and a single year for a 20-account firm tells you little. Track the three measures for several periods before drawing a conclusion.
Now the cost side. Take one hypothetical account priced at $1,800 a month with a gross margin of 30 percent. Monthly gross profit is $1,800 × 0.30 = $540, or $6,480 a year. A quarterly review that takes 1.5 supervisor hours, at an assumed fully loaded cost of $28 an hour, costs 1.5 × $28 = $42, so four reviews cost $168 a year, which is 2.59 percent of the account's annual gross profit. Divide $6,480 by $168 and you get 38.57. If a review programme saved one account of this size in a year, it would cover the cost of reviewing up to 38 such accounts (38 × $168 = $6,384; 39 × $168 = $6,552 would exceed it). The calculation ignores the cost of winning a replacement and assumes the reviews save an account, which we have no evidence for. Its purpose is to show scale: review time is small next to what an account earns.
For margin and cash across a whole company, see cleaning business profit margins and cash flow. Staff stability affects consistency at a site, and hiring, training and retaining employees covers that side.
When an account leaves anyway
Some accounts go regardless of service: a building sells, a company moves, a new owner brings a preferred vendor. Others leave over something you could have changed, and the exit conversation tells the two apart. Ask the departing contact, politely and once, what led to the decision and what a vendor should do differently, and accept the answer without arguing. Record it in the account log with the final margin and the last year of inspection history. Look for patterns after three or four losses, not one. Stay courteous, since people change employers and buildings change hands, and close out the final invoice and the return of keys and access cards.
Losing an account also opens a hole in monthly revenue that a pipeline has to fill, and pipelines take time to build. Good Loux, which publishes this Journal, makes software for finding, researching and contacting commercial prospects. It does not schedule crews, run inspections, score quality or report to clients, and it cannot keep an account for you. The demo shows what it does. Our templates article covers the follow-up discipline, and how to scale a cleaning business discusses the capacity new accounts require.
A first 90 days of a retention routine
This plan covers actions within your control. It does not promise that any client will stay.
Days 1 to 30. Pull every contract and write the end date, notice terms and scope summary into one list. Name an owner for each account. Build the scorecard for your three largest accounts by editing the line items to match each scope. Do a first scored inspection of each. Start a complaint and communication log.
Days 31 to 60. Inspect the three accounts again and compare. Send each client contact the result and a short plan for open items. Hold a first quarterly-style review with the largest account. Add the nine warning signs to your review notes. For any contract ending within 180 days, start the renewal worksheet: scores, complaint log, hours against budget.
Days 61 to 90. Extend the scorecard to the remaining accounts. Calculate account retention, gross revenue retention and net revenue retention from your month-end list. Put the 180-day and 120-day dates for every contract on a calendar. Decide your policy on price increases and on notice, and have an attorney review the contract language that governs them.
Sources
We read these pages directly on November 5, 2026. Pages change, so check the originals. BSCAI is a trade association, and several of its articles are written by contractors or consultants describing their own experience. They are cited for practice, not as measurements.
- Contracting Profits and BSCAI, 2024 Report on the Building Service Contractor Market (May 2024) for the survey method, account turnover distribution, average accounts per respondent and customer priorities.
- ISSA, CIMS: The Cleaning Industry Management Standard, 2023 edition for the quality-measurement, inspection, complaint and feedback provisions and the definitions of "shall," "should" and "may."
- BSCAI, 5 Tips for Managing Contract Cleaning Customer Relationships (March 3, 2026).
- BSCAI, From Chaos to Consistency: Turning Customer Complaints Into Retention Gold, Nicole Bernardo (November 5, 2025).
- BSCAI, Preparing for Renewals in the Janitorial Contracting Space: Strategy Over Scramble, Kristy Elmore (September 2, 2025).
- BSCAI, 9 Scary Signs Your Account May Be Going Out to Bid, Kristina Thayer (October 27, 2026).
- CleanLink, Implementing A Cleaning Quality Control Program, Mickey Crowe (January 29, 2015).
The scorecard, its weights and bands, the suggested inspection schedule, the renewal calendar's later dates and every worked example are the Journal's own and are not research findings or benchmarks. Nobody was interviewed, and no figure in the worked examples describes a real company. This is general education, not legal or employment advice: have an attorney review contract terms, and check your state labor agency on employment questions.


