How commercial cleaning firms find clusters of offices to win
A cleaning contract is easier to win and cheaper to run when the next client is in the same tower or street. How to find office clusters and the person who signs.
Part of the guide Workflow automation
Commercial cleaning firms grow fastest by winning offices close to the ones they already clean: the same tower, the same business park, the same few streets. Find those clusters by listing the offices and business centres in your city, counting how many sit within a short walk of each other and of your current sites, and then finding the facility or office manager at the busiest ones. A crew that walks between three clients in one building costs far less than one that drives across town.
This is for owners and sales leads of office cleaning, washroom service, window cleaning and facility service firms, from a five-van business to a regional company. It explains why density matters more than size, how to read a cluster, a made-up week in Johannesburg, the traps that fill a pipeline with offices that will never switch, and an AI employee that refreshes the list every month.
Why a cluster beats a big name
A large headquarters is a tempting target, but it usually has a facilities team, a tender and a contract that runs for years. A cluster of small and mid-sized offices works differently. Each one decides quickly, often through one office manager, and each extra client in the same building adds almost nothing to your travel, parking or supervision. Your night supervisor checks four floors instead of four suburbs.
Clusters also sell themselves. "We already clean the third and fifth floors" is a reason to switch that no brochure can match, and a building manager who sees your crew every night may recommend you to the next tenant.
How to find the clusters
- List the office-type places in your area with Nearby places: corporate offices, business centres, coworking spaces, and the professional firms that rent floors, such as accountants, law firms and clinics.
- Read the building. Many places in the database say which building or mall they sit in, and on which floor, when it is known. Offices that share a building name are one cluster.
- Count the neighbours. A second Nearby places block, set to pass on each office with a count, says how many other offices sit within 200 m of it. A high count marks a dense street or business park.
- Bring in your own sites from My data and count the prospects within a short walk of each, so the list starts with the offices nearest the crews you already run.
- Rank and write. xMap LLM can score each office as a cleaning prospect from what the data says (its kind, its rating and number of reviews, whether it lists a website, how many offices surround it) and write one line on why.
Reading the list well
The places database tells you where offices are and what kind they are. It does not tell you who cleans them now, when their contract ends, or how many square metres they rent. Treat the list as a map of where to knock, not a forecast of who will buy.
- Building name missing does not mean the office is alone. Many listings carry only an address; offices at the same street number usually share a building.
- A coworking space is one client with many tenants. Its operator usually buys cleaning for the whole space.
- Ratings and reviews say how busy and visible a place is, not how dirty it is. Do not read a low rating as "needs better cleaning".
- Clinics and labs often need specialist cleaning with their own rules. Keep them in a separate list if you do not offer it.
A worked example (made up)
Say a cleaning firm in Johannesburg looks after fourteen offices in Sandton and Rosebank. Once a month its AI employee lists every corporate office, business centre and coworking space within 2 km of those fourteen sites, counts the offices within 200 m of each, and keeps the ones in buildings where the firm already works or with at least five office neighbours.
The list that arrives might hold forty offices. Nine share two towers the firm already cleans; the rest sit in three office parks. The sales lead calls the nine first, opening with the floors the crew already covers, and books walk-throughs at the office parks for the following week. These numbers are invented to show the shape of the work, not results from any firm.
Mistakes that fill the pipeline with the wrong offices
- Chasing the biggest names. Head offices buy through tenders. Unless you are ready for one, start with the floors around them.
- Ignoring travel. An office 15 km from any crew is a loss-maker on a nightly contract. Count the distance to your nearest site before you price.
- Calling the wrong person. The tenant's office manager signs for the office; the building manager signs for shared areas and windows. Know which job you are selling.
- Sending the same list twice. A monthly list without memory repeats last month's offices. Let the flow remember what it already found.
- Writing to everyone at once. A first message that names the building and the crew already there works; a generic one does not. Approve each one before it goes.
Doing it with an AI employee
An AI employee in xMap Flows can rebuild the cluster list every month and hand the sales lead only what is new. Its blocks, in order:
- Schedule: the first Monday of each month, 07:00.
- My data: your current client sites.
- Nearby places: offices, business centres and coworking spaces within 2 km of each site.
- Nearby places again, set to counts: how many offices surround each one within 200 m.
- Only new: offices this flow has not found before.
- xMap LLM: a fit score and a one-line reason for each office.
- Sort and limit: the 30 highest scores.
- Apollo: the office or facility manager at each company, from your own Apollo account.
- Email and Show on a map: the list and the clusters on a map.
What lands in the inbox each month is a short ranked list: each office with its building, how many neighbours it has, the nearest site you already clean, its fit score and reason, and the person to ask for. To write first messages, add Outreach, which waits for your approval of each one. The template Office towers for window and facade cleaning is a ready start, and New apartment buildings for cleaning contracts does the same for residential blocks. See also finding businesses that share a building.
Questions
Can it tell me who cleans an office now?
No. The data says where offices are and what kind they are, not who their suppliers are or when contracts end. Ask on the first call.
How close should offices be to count as a cluster?
Within a short walk, so one crew or supervisor can cover them in a night. 200 m is a good first setting; widen it for business parks with car access.
Does every office show its building?
No. The building and floor are there only when known. Offices at the same street number usually share a building, and the AI step can group them by address.
Will it send messages to offices on its own?
Only if you set it to. Outreach waits for your approval of each first message, and you can rewrite it before it goes.