See it working · Case studies

Small levers, big shifts.

Real results from our founding partner's own businesses: what changed when we put the lever in the right place. The businesses aren't named, the work and the numbers are real.

Cleaning Security Websites Group

Case study 1 · Cleaning · 750+ staff · about 150 sites

A cleaning company that joined up its information and gave its managers time back

A regional cleaning company with 750+ staff across about 150 sites, mostly schools, brought information from half a dozen places into one operations platform, and now wins tenders with a standard, evidence-led approach.

The situation

Thirty years of steady growth meant information had built up wherever it was easiest at the time: SharePoint, shared drives, Google Sheets, email and Excel. It worked, but pulling together a site file, a tender or a monthly client report meant knowing where everything was and who to ask.

Where we put the lever

Mapped what was where: every list, folder and spreadsheet, how current it was and who used it, so only what's live moves forward.
Brought it together in one platform: each area manager sees just their own sites, with the site file, quality audits, actions, staff and DBS register in one place, and schools grouped under their trust.
One format for everything: tender packs, zoned cleaning plans, job cards and site documents now follow one standard, so what's written for a bid becomes the working plan once the contract is won.
Learned from every bid: evaluator feedback is turned into rules for the next tender, so each bid starts from the last one's best answers.

How far it tipped the balance

A secondary school contract won back with 91.74%, including full marks on term-time cleaning, deep cleaning, staff cover, mobilisation and value for money.
The tender pack that scored full marks is now the standard for every school bid.
The platform is being piloted on the newly won school, with roll-out across sites planned from February 2027.

What the team can do now

The balance has tipped back: managers spend less time finding information and more time with their sites and clients, and the company can show schools and trusts exactly what they're getting, school by school.

OneView demo Long Lever Academy guide

Case study 2 · Security · guarding, patrols, keyholding, 24/7 control room

A security company that gave every officer the same clear instructions

A regional security company running guarding, mobile patrols, keyholding and a 24/7 control room standardised the instructions behind every site and moved patrol records off paper, so officers and the control room work from one consistent picture.

The situation

Each keyholding site had built up its own instructions over the years, written by different people in different formats and kept as long documents. Patrol runs had been recorded on spreadsheets and paper. Everything was there, but finding the one line an officer needed on site took longer than it should.

Where we put the lever

One standard for every site: keyholding and alarm-response instructions rebuilt into 20 templated procedures, so every new site is set up the same way and every officer knows where to look.
Patrol records on mobile: patrol runs, alarm activations, incidents and supervisory checks recorded as they happen, searchable and reportable instead of filed.
Next: bringing site instructions, risk assessments and procedures together in one platform, and an assistant that lets the responding officer ask a simple question about that one site and get the answer straight from its instructions.

How far it tipped the balance

Every new keyholding site is set up from the same 20 templated procedures.
Patrol and incident records are searchable and reportable for clients, instead of sitting on paper and spreadsheets.

What the team can do now

One small change, a long way moved: supervisors and managers keep procedures current in one place, and clients get consistent service and consistent evidence, whichever officer attends.

Field Reporting demo

Case study 3 · Websites · SEO and GEO

Four websites rebuilt in-house to be found by search engines and AI assistants

A group of service businesses rebuilt its four websites itself (group, security, fire and security systems, and cleaning), set them up for Google and AI search, and turned them from brochures into tools that answer customers' questions and bring in enquiries and applicants.

The situation

The old sites had been built some years earlier on an older platform. They described services rather than answering the questions customers search for, old pages and duplicate addresses were still being picked up by search engines, and vacancies went out through job boards.

Why it matters now: the shift to AI search. More and more buyers no longer type a few words into Google and click through ten links. They ask an AI assistant such as ChatGPT, Gemini, Copilot or Perplexity a full question, like “who provides keyholding and alarm response near Hull?” or “how should a school score a cleaning tender?”, and get one answer that names a handful of businesses. Google now puts its own AI summaries above the links too. If your site isn't one the AI trusts and quotes, you can be left out of the answer altogether, however good your service is.

Getting into those answers is called GEO, generative engine optimisation: the AI-search partner of SEO. SEO helps you rank in the list of links; GEO helps AI assistants understand who you are, what you do and where, and trust you enough to recommend you. The two reinforce each other, and done well they build what both reward most: authority, meaning a site that clearly and consistently answers real questions, backed by evidence.

Where we put the lever

Rebuilt each site around the customer's question: service hubs with detailed sub-pages, and location pages for each town served, written around local searches and local teams.
Made them easy for search engines and AI to read (SEO and GEO): a title and description for every page; structured data that tells search engines and AI exactly who the business is, what it does, where it works and which questions each page answers; sitemaps; and pages served in a form AI crawlers can read.
Written to be quoted: pages and guides that each answer one real customer question in plain English, answer first, with UK specifics and sources: the kind of content AI assistants lift into their answers.
The same story everywhere: one name, one description, one set of contact details and accreditations across every site and listing, so search engines and AI join the dots and treat the group as one trusted source.
Cleaned up the past: old addresses redirected properly and duplicate pages removed, so search engines put their weight behind one page per topic.
Tools that do the first conversation: cleaning hours and specification tools, a security decision-support tool, and resource libraries with guides that answer questions and quote their sources.
Careers on their own sites: every vacancy on its own page, ready for Google's job search, with quick apply and a reply to every applicant.

What is llms.txt, and what we did with it. An llms.txt file is a short, plain-text summary that sits on a website at /llms.txt, written for AI assistants rather than people. Where a normal web page is full of menus, styling and code, llms.txt tells an AI in a few hundred words exactly who the business is, what it does, who it serves, where it works, what makes it different, and which pages to read for each topic, with links. It's a newer, proposed standard: not every AI assistant has confirmed it reads the file, but it costs nothing to add and gives any AI that does a clear, accurate account in the business's own words, rather than leaving it to guess from scattered pages.

For this website we wrote our own by hand rather than letting the website platform generate a generic one: a one-paragraph summary, the points of difference, who we help, how we work and what it costs, the key pages and articles with a one-line description of each, and the common questions we answer. We paired it with a robots.txt file that explicitly welcomes the main AI crawlers to read the public pages, and a sitemap listing every page, so nothing important is missed.

How far it tipped the balance

In one 90-day period, four of the security company's town pages alone drew 5,276 Google impressions.
24 vacancies indexed in Google's job search from the cleaning company's own site, with no job board fees.
Duplicate-content warnings cleared across the security site's main service pages after the redirect clean-up.

What the team can do now

The weight has shifted from paying others to doing it themselves: the sites are changed in-house, the same day, by people who know the business: new pages, new vacancies, seasonal offers and updated guides, without paying anyone for each edit. The same method built this website.

Long Lever Interiors marketing guide Paying an agency for SEO and AI search?

Case study 4 · Group · policies, procedures and documents

A £25m+ group that brought its paperwork up to date in months

Across a group of service businesses, policies, procedures and documents were reviewed, refreshed and brought into one format and one place, so staff always find the current version and new requirements are covered before anyone asks.

The situation

As each business grew, policies and procedures were written and updated at different times, by different people, in different styles, and saved on SharePoint, shared drives, email and Excel. Some had been overtaken by new rules and ways of working. Finding the current version, and showing it to a client, an auditor or an insurer, took more effort than it should.

Where we put the lever

Took stock: listed every policy, procedure and template, where it lived, when it was last reviewed and who owned it.
Refreshed what was out of date and retired what was no longer needed, then put everything into one consistent format and one brand.
Filled gaps early: for example, an AI and data use policy and a staff guide, with sign-off tracked, before insurers asked for them.
One home, current versions only: a single compliance library, with each business's policies kept separately and linked into the operations platform; only live documents carried forward, the archive left behind.

How far it tipped the balance

One format and one library for policies and procedures across the group.
An AI and data use policy and staff guide in place, with sign-off tracked.
Only current documents carried forward into the new systems.

What the team can do now

Hours that went on hunting for documents now go on running the business: every manager works from the same current version, reviews happen on a schedule rather than when someone remembers, and bids and audits draw on a library that's already up to date.

Safe by design

Where would the lever go in your business?

The free discovery check takes a few minutes and points to the first, best place to start. We guide; your team builds it and owns it.

Take the free discovery check