Get More Leads as a Builder: A 30-Day Plan Drafted With AI

A small builder can get more leads in 30 days without a directory by working four weekly blocks: complete the Google Business Profile, ask every past client for a review and an introduction, publish three genuine project pages, then turn that proof into posts and review replies. Log where every enquiry came from.

None of this is instant. A stronger profile and fresh reviews work on people who are choosing a builder right now, but the effect shows over weeks, so judge the month against your own baseline: the enquiries you logged in the three months before. AI drafts the wording. You supply the real jobs, photos and client permissions, because the risk is a confident page about work you never did.

Follow me on Instagram@sagnikteaches

Before day one: a fact sheet AI is not allowed to improve

Everything in the month rests on one private document. Write a fact sheet for the firm: trading name, the services you really do (and the ones you turn down), the towns you cover, the true number of years trading, insurance and accreditations exactly as printed on the certificates, typical job sizes, team size, opening hours and how fast you promise to reply. Add five finished projects with one line each: what it was, how long it took, and one thing that went awkwardly and how you dealt with it.

Connect on LinkedInSagnik Bhattacharya

The example running through this tutorial is a five-person firm doing extensions and loft conversions. It has 46 past clients from the last four years, 11 Google reviews, and nine enquiries logged in the past 90 days, about three a month. Those last two numbers are the baseline. Without them, week four is just a feeling. Paste the fact sheet at the top of every prompt below, because the AI may draw on nothing else.

Subscribe on YouTube@sagnikteaches

Use ONLY the facts between FACTS and END FACTS. If you need a detail
that is not there, write [MISSING: what you need] instead of guessing.
Never add awards, accreditations, years in business, prices, place
names, client names or numbers that are not in the facts.
British English, plain words, no superlatives.

FACTS
Trading name: [name]
Services: loft conversions, rear and side extensions, kitchen refurbishments
We do not do: roofing-only jobs, new builds
Towns covered: [Town A], [Town B], [Town C]
Years trading: [true number]
Team: five people, two vans
Reply promise: same working day
Projects: (1) loft conversion, six weeks, steel beam over the stairs...
          (2) rear extension, ten weeks, ...
END FACTS

Week 1: make the Google profile the strongest page you own

Your profile shows in Maps and Search next to your name, and unlike a website it costs nothing to improve. Do the plain jobs first. Claim and verify it, because Google only shows performance data for verified profiles. Pick the closest primary category Google offers for your main work and add secondary categories for the rest. Check that your phone number is on the profile: Google's call count only works once a number has been added there.

1. Add every service you offer as its own entry, with a short description (the first AI job below).
2. Write the business description (the second AI job).
3. Upload at least ten real photos of finished work, with the permission of the clients whose homes appear. Real photos only: an AI-made picture of a loft you never built is a false claim about your work.
4. Schedule one post a week using the profile's own scheduling option. Posts older than six months are archived unless dated, and events need start and end dates, so don't treat posts as permanent storage.
5. Copy the link that opens your review box directly, and keep it in your phone notes for week two.

Now the AI jobs. Paste the fact sheet block from above, then add this request.

Write: (1) a profile description of about 120 words in three short
paragraphs; (2) one service entry for each service in the facts, two
sentences each, saying what the work includes and what it does not;
(3) six post ideas, each tied to one of the projects in the facts.
Flag anything you had to assume with [CHECK].

What comes back usually resembles the draft below (illustrative). Read it as a proofreader hunting for claims, not for style.

We are a five-person building firm that designs and builds loft
conversions, rear and side extensions and kitchen refurbishments for
homeowners across [Town A], [Town B] and [Town C].

Every job has a named site lead who is on site daily. We give a fixed
written price after a survey and explain what is and is not included
before work starts.

Award-winning craftsmanship, fully accredited by all leading trade
bodies. Call today for a free, no-obligation quote.

The first paragraph is safe because it restates the fact sheet. The second paragraph invents a site-lead system and a pricing policy the sheet never mentioned, and the third borrows two claims from the AI's idea of a typical builder. Delete "award-winning" and "fully accredited" unless you hold the award and can name the body. Keep the site-lead sentence only if it is true on every job, and keep "fixed written price" only if you really quote that way. If you do not offer free quotes, take that out as well, because a line you cannot honour is a bad review waiting to happen. For the rest of the profile settings, see setting up every part of a Google Business Profile with AI.

Service entries need the same read. One illustrative draft for loft conversions came back as "Complete loft conversions, all approvals handled, finished in as little as four weeks." The fact sheet gives six weeks for the one loft it lists and says nothing about who deals with approvals, so the entry becomes: "Dormer loft conversions, from survey to finished room. Our last one took six weeks on site, after a three-week wait for the engineer's calculations. We do not take roofing-only jobs." It is less punchy, but every clause can be checked against a real job, and the waiting time warns a buyer about the part that frustrates most people.

Week 2: ask every past client, with one personal message each

Reviews and introductions from people you have already built for cost very little, because the only price is a little awkwardness. Export your past clients from your invoicing or job software into a sheet with first name, job type, completion month and the best way to reach them. In the example firm that gives 46 rows. Remove anyone you cannot reach or who has an open dispute, which in this case leaves 38. Send the requests in batches of ten, so you can answer replies properly.

The old way is one pasted line. The better version names the job, gives a reason, and keeps the ask small. Both are shown here, with placeholders where the real names go.

BEFORE
Hi, we'd love a Google review! Link here. Thanks.

AFTER
Hi [first name], it's [your name] from [firm]. I was thinking about
the loft conversion we finished for you last spring and hoped it is
still working well. If you have two minutes, a short Google review of
how the job went would help other homeowners choosing a builder.
Honest is what helps, good or bad: [review link]. If anything about
the job still isn't right, tell me directly as well and I will fix it.

Three rules sit behind that wording. First, send the same request to everyone, whatever you expect them to say: Google's review rules rule out selectively asking for positive reviews, and its help page is plain that offering free or discounted work in exchange for a review is prohibited. Second, never mention a discount, gift or prize in the same message. Third, keep referral requests separate, in a second message a few days later, and never link a thank-you for an introduction to a review being posted.

AI earns its keep on the variants. Ask it for three versions of the message for a loft client, an extension client and a kitchen client, each using one real detail from the project line in your facts, with [first name] and [your name] left as placeholders. Then read each one aloud. If it sounds like something you would not say on a doorstep, change it.

A quick sum shows what to expect. Of 38 people asked, suppose one in four posts a review. That is nine or ten new reviews, taking the profile from 11 to about 20. One in four is an assumption, not a benchmark, and your rate could be lower. If one in ten of the 38 replies with a name for you to call, that is four introductions, which for a firm with three enquiries a month would be a large change.

The same week suits a second, offline job. When a project is under way, ask the client's permission and put a short note through the neighbours' doors, giving the working hours, a named contact and a line saying you build extensions and loft conversions nearby. Ask AI to draft it from three facts: working hours, contact name, and what you are building.

Week 3: three project pages, not thirty town pages

This is where AI-written local content goes wrong. The temptation is to generate thirty near-identical pages called "Builders in [Town]". Google's spam policies name this pattern: its section on doorway abuse names pages aimed at particular regions or cities that funnel visitors to one page, as well as substantially similar pages. Read the wording yourself on Google's spam policies page before you publish a batch of anything.

Write three pages about real finished projects instead. Each page covers the job type, the awkward part and how you handled it, the timeline, six captioned photos, one client quote with permission, and a plain contact button. Dictate the story into your phone's keyboard on the drive home, since that is faster than typing and keeps your own phrasing. Then paste the transcript under the fact sheet.

Write a project page for the loft conversion in my notes below.
Headings: The brief, The awkward part, What we did, How long it took.
Under 350 words. Use only the notes. Put [PHOTO: what it shows]
where a picture belongs. Do not describe the street, the
neighbourhood or the client.

NOTES
Dormer loft, three-bed house, steel beam over the stairs, engineer's
calculations took three weeks, build six weeks, client lived in
throughout.

A typical draft (illustrative) will call the house "a charming period property close to the town centre", say the work "was completed ahead of schedule", and add a line about "premium Scandinavian insulation". None of that was in the notes. Strike the location detail, which you cannot verify and the client may not want published, and strike "ahead of schedule" unless you have the programme to prove it. If you did use a named insulation product, write it in yourself. Keep what the AI does well, which is structure: short headings, active sentences, and a first paragraph that says what the project was.

Link each page to the next and to a single "get a quote" page, and put your service towns in a plain sentence on each one ("we work in [Town A] and [Town B]") instead of making a page per town. Once the pages are live, add each link to a profile post.

Week 4: turn the proof into posts, replies and one case study

By now you should have new reviews, three project pages and a stack of photos. Week four is about showing them in the places buyers look. Reply to every review, new and old, in two or three sentences that mention something specific. Google's own guidance says to avoid sending the same "thank you" to everyone and to keep replies short and conversational. Here is a reply to a four-star review that mentioned a slow start, written from a prompt that supplied the review and the facts (illustrative).

Thank you, [first name]. You are right that the first fortnight was
slower than we had planned, because we were waiting on the structural
engineer's calculations before the steel could be ordered. I would
rather tell you that at the time, and we are changing how we update
clients during a wait like that. Glad the finished loft is working
for the family. [your name]

The reply avoids three traps: it does not argue, offer a discount or promise anything the firm has not decided. If AI proposes an apology for something that was not the firm's fault, remove it. For a bigger set of examples, including negative reviews, see replying to good and bad Google reviews with AI.

Next, one case study. Pick the project with the best story and ask the client three questions by message: what worried you at the start, what surprised you, and what would you tell a neighbour. Their words go in quotes, unedited apart from typos, and the case study method using customer interviews shows how to turn answers into a page without inventing quotes.

Finally, spread the material across the week with a simple pattern: a finished-job photo with two sentences on Monday, a before-and-after pair on Wednesday, a client quote on Friday. Ask AI for three captions per photo and pick the plainest. A caption like "Dormer loft, three-bed house, six weeks on site" beats three lines of adjectives.

A one-sheet lead log that shows what is working

Create a sheet with these columns: date, first name, how they found you, job type, budget they mentioned, minutes until your first reply, site visit booked, quote sent, outcome. Fill it in the moment an enquiry arrives, even if it is a call that goes nowhere. Ask every caller "How did you hear about us?" and write the answer in their words. Google's profile numbers help, but they do not catch everything: the call count only records taps on the profile's call button, so a caller who copied your number from a van or your website will not appear in it.

Each month, also note the profile's own figures for calls, website clicks, directions requests and messages. Google lists them on its performance insights help page, and notes that a profile shows only the metrics relevant to that type of business. Below is the sheet at day 30 for the example firm (illustrative figures).

Twelve enquiries against a baseline of three a month is a big jump, so treat it with suspicion. Part of it may be pent-up demand from earlier, and a single month is a small sample. The table's useful message is in the site-visit column: introductions converted best, so the week-two asks deserve repeating every quarter, while the single social enquiry says to keep posting but not to spend more evenings there. Check the reply times as well. If your slowest replies were also your lost enquiries, a reply routine matters more than any extra page, and a routine for answering trade enquiries within a minute covers that.

What the month costs in cash and hours

The plan can run for nothing, which is the point. Here is where money is optional, using list prices checked in late October 2026. Vendors change prices and some hide them behind sales forms, so confirm on the vendor page before buying.

The sensible middle version is the free AI tier, Canva Free and your existing website, which costs $0. If you need a site and want the paid assistant, it is about $39 a month. The version with every paid row comes to $176 a month, which is rarely justified in month one.

Time is the real cost. Budget about four hours in week one, three in week two, five in week three and three in week four, which is 15 hours of setup, plus roughly 10 hours across the month for the extra replies and logging. Call it 25 hours. At $45 an hour for your own time that is $1,125, so 12 enquiries would cost about $94 each with no cash spent. That is the number to hold against whatever a lead directory charges per lead, which you can read straight from your invoices, and how much Bark charges for leads shows how to work it out. In month two the setup hours disappear. If upkeep takes 10 hours and the same 12 enquiries arrive, the figure drops to about $38 each.

How the plan goes wrong, and the checks that catch it

Five things can derail the plan. The first is the invented claim: an AI draft says "fully insured up to $10 million" because it sounds standard. The check is to compare every number and every named body in a draft against a printed certificate, and delete the rest. The second is the incentive, where a friendly offer of "10% off your next job if you review us" turns a lawful request into a prohibited one. The third is the duplicated page, where only the town name changes. The fourth is slow follow-up, where the leads come but sit for two days. The fifth is stopping on day 31.

Here is a realistic example of the first problem (illustrative): a firm asks for a service-page draft with no fact sheet, and the AI adds "our team of chartered structural engineers". The firm employs none. A client who reads the page asks for one to attend a meeting, which turns a line of copy into an awkward admission. With a fact sheet in the prompt, the same request would have returned a [MISSING] marker instead.

To check the month worked, answer four questions on day 30. Did the profile get its ten photos and weekly post? Did you ask every one of the contactable past clients, not just the easy ones? Are three project pages live, each with real photos and no invented detail? Is every enquiry in the log with a source? If all four are yes, the numbers will tell you what to repeat. On day 31, put a recurring 30-minute slot in your diary for the lead log and one post a week, and keep the week-two ask running as each job finishes, so the next reviews arrive without a campaign.

Questions builders ask before starting the 30 days

How many extra enquiries should a small builder expect in 30 days?

Nobody can promise a number, and anyone who does is guessing. Count the enquiries you logged in the previous 90 days, divide by three, and treat that as your baseline. Set a target you would be happy with, such as one or two more a month, and judge the plan by your own sheet at day 30 and again at day 60, because reviews and pages keep working after the month ends.

Can I offer past clients a discount for leaving a review?

No. Google's help page on getting more reviews says offering incentives, like free or discounted goods or services, in exchange for reviews is strictly prohibited, and its content policy removes incentivised reviews from Maps. Ask every past client the same way, whatever you expect them to say, and keep any referral thank-you completely separate from the review request.

Do I need a website to follow this plan?

The Google profile and the review asks work without one. Week three needs somewhere to publish project pages, so if you have no site, a one-page website on a basic plan is enough; the cheapest Squarespace plan lists at $19 a month on annual billing. Without a site, post the same project stories on the profile instead.

Which AI tool should I use for the drafts?

Any mainstream assistant will do, including the free tiers of ChatGPT, Claude or Gemini. What matters is that you paste your fact sheet first and tell the tool to use only those facts. A paid plan at about $20 a month is worth it only if you will use it for quotes, emails and site notes as well as this plan.

Further reads

• How to Write Google Business Profile Posts With AI Every Week — A weekly posting routine once the profile is set up.
• What AI Can and Can't Do for Your Local SEO — What AI can and cannot do for local search.
• How to Run a Customer Referral Programme With Automated Reminders — Keep introductions coming after the first month.
• Can You Use AI to Write Testimonials? Where the Legal Line Is — Where editing a client's words crosses a line.
• AI Marketing for Contractors: Reviews, Case Studies and Local Posts — A wider marketing routine for contractors.
• The First 30 Days of AI in a Trades Business — AI jobs beyond marketing for the same month.
• Deck Builder Leads: What They Cost and How to Reply Fast With AI — Work out the most a deck lead can cost you, compare Thumbtack, Bark, Angi and your own site by cost per booked visit, and reply in minutes with AI drafts.
• Best AI for Builders: Pick the First Job by Hours Lost and Risk — A scoring sheet for small building firms: rank missed calls, quote follow-up, payment chasing and site notes by hours lost and risk, then start with the winner.
• Best AI for Marketing Ideas: A 12-Month Plan for a Garden Room Builder — Twelve worked marketing ideas for a garden room builder, one per month, with prompts, checks and a number to watch, built into a calendar with AI.
• ChatGPT for Builders: Ten Prompts for Quotes, Chasers and Client Emails — Ten copy-and-paste ChatGPT prompts for domestic builders, from scope notes to overdue chasers and handover emails, with sample replies and what to fix.
• Fence Company Lead Generation: Follow Up Every Enquiry With AI — A first reply that asks for run length and photos, a visit-booking routine, a follow-up schedule and the sums that show whether to buy more leads.
• Fence Company Marketing: Local Pages, Photos and Posts With AI — A claims sheet, town pages that pass Google's spam tests, a photo caption routine, a posting rhythm and a four-week plan a small crew can keep up.
• Get Landscaping Clients: Project Photos, Local Search and AI Follow-Up — Week-by-week steps for landscapers: tidy the Google profile, publish captioned project photos, request reviews and follow up every enquiry with AI drafts.
• Home Builder Lead Generation: Nurture Enquiries With an AI Sequence — Follow up website enquiries with six AI-drafted messages on plot, budget and timing, sent by Zapier or your CRM, with consent and opt-outs handled first.
• How to Price Building Work: Check Your Rates Against Past Jobs With AI — Compare quoted and actual cost line by line on past jobs, find what you underprice, and turn it into a rate sheet you test before quoting.
• AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.

Sources: Google Business Profile help pages on reviews and performance metrics, Google Search Central spam policies, and vendor pricing pages for ChatGPT, Claude, Canva, Squarespace, CallRail and NiceJob (all checked October to November 2026).