Build the marketing your firm needs, one part at a time.

Your website is your retail storefront.

Fix it before you send a single visitor. Traffic into a store that scares people away is money leaking out the front door.

Start here: search your own city. Your competitors set the bar, not me.

Do this right now, before you touch anything else: Google "structural engineer near me" in your city. Write down who comes up on the map and in the first results.

Those firms are your competition. Your website needs to be at least as good as theirs. That's the baseline, and it's set by your market.

In some cities I've searched, the sites at the top of Google are outdated and the bar is low. In others the market is competitive and the top sites are beautiful.

So the level of build you need is dictated by your competitors. Don't over-build for a market that doesn't demand it. Don't under-build for one that does.

Keep that list. You'll use it again for reviews, for service pages, and for deciding whether paid ads are cheap in your market.

The main strip: if the store scares people away, you're bleeding money.

Walk down the main strip of any city. You go into the restaurants that look beautiful, open and confident, not the ones that look outdated or crummy. Your website is that storefront.

You can send as much traffic and do as much marketing as you want. If people come into the store and the store scares them off, every dollar you spent getting them there leaks out.

That's why the storefront is step one. Traffic comes second, always. Nobody pays for a billboard pointing at a boarded-up shop.

Building a good site is not that difficult once you understand the basic elements. There are three, plus a bonus that matters more than people think.

The three elements of a website that does its job.

In plain English: design is how it looks, the journey is how a visitor becomes a contact, SEO is how Google finds it.

Design

It needs to be beautiful to some degree: a firm you'd trust with a building.

User-friendly: an easy, fluid path from the homepage, through the service pages, to the contact page.

Judge it against the sites from your own search, not against a template.

The journey (CRO)

Conversion rate optimization: turning a visitor into a contact.

You can have a beautiful website, but if it doesn't capture contact details, what is the point? To make money you need the details, so the relationship can start.

Homepage, then the services you provide, then a clear reason to reach out.

The more time they spend reading, the more they qualify themselves, and the less time they spend on a competitor's site.

SEO

Search engine optimization: being found on Google.

The right keywords and the technical structure Google expects.

Google's stated aim is to give the searcher the most relevant, useful result. Building by its guidelines makes a page eligible to be found; Google says plainly that it does not guarantee a page will be crawled, indexed or shown, even when the page follows them.

A lot of the sites I'm shown were built without this part. It's where the attention comes from.

The contact form: ask the real questions on the front end.

I still see a lot of structural firms with a form that is a name, an email and a little message box. That form qualifies nobody and saves nobody time.

Put five, six, seven real questions on it. You ask the same five or six questions on every first call anyway, so you might as well get them answered before the call.

It qualifies the person before you spend an hour with them, and it speeds up the sales cycle: the deep conversation happens sooner because you already know the project.

Which questions? The ones you ask on every first call. Write that list down this week. That list is your form.

The journey side of the same idea: you want people to spend time on your site, so they educate and qualify themselves before they reach out, and they're not spending that time on your competitors.

The bonus element: speed. Ship a B-plus in two weeks, not an A-plus in six months.

I see entrepreneurs spend way too much time on perfection. I'd rather ship a site that scores a B-minus to B-plus on design, journey and SEO in a week or two than an A-plus on all three that takes six or seven months.

The opportunity cost is real. Every inquiry that happened in those months went to somebody else. The gap between one week and six months is a cost of the launch, and nobody puts it on the invoice.

Speed is a competitive edge. Once the site is bringing leads and revenue, put money back into it and take every element from B-minus to A-minus to A-plus.

Speed beats design. Even the journey beats design. I don't care to build pretty websites. I care that the asset does its job, which is acquiring leads. A beautiful website that does nothing is a beautiful website that does nothing.

The client who waited two months for a website, and then nothing happened.

Before working with me, one of my clients had been through two marketing agencies. The most recent one took two months to build his website. After it launched, nothing happened.

We built his in four days. His words, not mine, on a recorded call.

The lesson isn't that four days is the target. It's that knowing the strategy is only half of it; you have to move. In the age of AI it's possible to move extremely quickly.

Every moment an asset isn't live and converting is potential revenue you're not collecting. A site that doesn't get attention, a site that doesn't capture leads, is losing you money while it waits.

Marketing you can see, or marketing you're guessing at.

Three free accounts. Many of the firms I meet have never set them up. Set them up the week the site goes live, before you spend a dollar on traffic.

Why I work off data, and why you should too.

Many of the companies I meet have no tracking. The power of digital marketing is that you can track everything: which channel brought the person, what they did, whether they called.

I would rather work off data than sell anyone a dream. I love to be right, but I'm willing to be proven wrong and pivot once I see the numbers.

It's so much easier to grow when you clearly see the inputs, the process and the output. If a number can't be verified, I'd rather say so than show it.

All my clients have access to their own data. They can see whether their business is actually improving. They don't have to take my word for it, and you shouldn't take anyone's.

The three free accounts, and what each one tells you.

Google Search Console

Am I getting attention?

Shows how often Google presents your website for particular keywords (impressions) and how often someone clicks.

If that curve isn't growing, you're not getting attention.

This is the data you use to fine-tune your SEO.

Google Analytics

Where did they come from, what did they do?

Every source of traffic: organic, paid, AI, referrals, social media.

How people interact with the site: how long they stay, how many pages they touch.

This is the data you use to improve the journey, so you can invest more in the channel that is doing more than you are paying it for.

Google Tag Manager

The one most people miss.

Wires Analytics, Search Console, call tracking and, later, paid ads on Google and Meta together, cohesively.

Lets you track specific campaigns and channels, so you know where the payoff is actually coming from.

Set it up once, at build, not bolted on later.

Read your own numbers: what to look at, in order.

Search Console, Performance report. Two lines: impressions (how often you appeared) and clicks (how often someone chose you). Look at the shape over months, not one day. Climbing month after month is the signal; a spike that dies is not.

Search Console, Queries. Which searches you appeared for. That list tells you which service pages to write next, because Google is already trying to show you for them.

Analytics, the channel report. Where the visitors came from and what they did. This is where you find the channel that is doing more than you are paying it for and put more into it.

Analytics, the pages. Where people stay and where they leave. If the contact page loses everyone, that's a journey problem, not a traffic problem.

Check yourself: do you have all three accounts? Do you know your impressions for last month? If the answer to either is no, that's the first thing to fix this week, and it costs nothing.

Three places, in the order they pay.

The map pack, your Google profile, and a page for every service you sell. Then the AI search results.

The map pack: three businesses, and the searcher needs one of them today.

Under the ads, Google shows a map with three businesses. That's the local map pack, and it's where the urgent homeowner searches go.

Nobody searches "plumber near me" because they're planning for a leak next year. They have a leak in the kitchen right now. Structural is the same: a permit issue, a GC issue, a renovation, fire damage. They need an engineer now, and the firms on the map are the first ones they see.

In the accounts I've worked on, small urgent jobs such as load-bearing wall assessments have been the steadiest source of inquiries. They are not the best projects, and nothing guarantees a full calendar. Search "load-bearing wall removal near me" in your city: are you in the three?

Google says three things decide who gets those three spots: relevance, distance and prominence. In practice that is your location, your reviews and the profile itself. One slide each.

Variable 1: location. Google wants to give a local answer.

If I'm in Miami and search "structural engineer near me", Google will not show me someone in New York. Distance is one of its three signals: it favors firms near where I'm standing, among the ones whose profile fits the search, and it knows where I'm standing.

So represent the location and the service area you actually serve. Google has two models: a customer-facing office it shows on the map, or a service-area profile for a firm that travels to clients and has no walk-in office. Both are legitimate; pick the one that is true of your firm.

If you keep a customer-facing office, put it where the work is. A room you actually work from, the logo on the door, video verification: for a firm that receives clients, one of the cheapest levers in this training.

The conditions for a displayed office, from Google's own eligibility rules: the office has to be staffed during the hours you list, clients have to be able to come to it, and it needs permanent signage. A mailbox, a virtual address, or a coworking desk nobody staffs does not qualify, and Google can suspend the listing. A firm that visits clients and has no walk-in office can still list as a service-area business; that is a separate, legitimate path.

Variable 2: reviews. The prominence you can build.

A lot of engineering firms never ask for a Google review. I've seen the top firms in serious cities sitting on 10 or 20 reviews, sometimes single digits.

Reviews feed the prominence part of Google's three signals: how many, how recent, how good. Steady new reviews from real clients are what move it; there is no published formula for the pace.

Do this: if you have a customer email list and you're a good service provider, go through every old client and ask. Then ask after every closed project, as part of closing it.

One, two, three a week is better than nothing. Some markets are competitive, but you have to start somewhere, and it compounds.

Variable 3: set the profile up correctly. Category first.

Your Google Business Profile is the listing that feeds the map. It's the profile that shows in the three, not your website.

The category comes first. Structural engineer. Civil engineer. If the category is wrong, it works against everything else: you are less likely to be shown for the right searches.

A client's story: a structural engineering client of mine had his profile in the general contractor category. We had to redo the Google verification, and it pushed every initiative back two to three weeks.

Then the rest: the office address you work from, hours, the services you actually sell, photos. Keep it accurate; Google checks.

Check yourself: open your profile now. Is the category right? Is the address the office you work from? How many reviews arrived in the last month? Three answers, three fixes.

The organic results: a page for every service you actually sell.

Below the map are the organic results. Google gives them to the website that has the right content for what the person typed. This is the nerdy part, and it's where most engineering firms miss.

Many of the engineering sites I see are a home page, a portfolio, an about page and a contact form. That's it. The right way is all of that, plus a page for every service and sub-service homeowners, general contractors and architects search for.

Structural design. Residential structural. Commercial structural. Load-bearing wall removal. Retaining walls. Foundation inspections. The list goes on. You can have 10, 20, 30, 50 service pages, one per real search.

It feeds the map pack too. Google reads your site to understand what you do. A profile that says structural engineer and a site that proves it is the combination.

A warning about AI content: pages a human would sign, or the work does not count.

Yes, we're in the AI era, and AI writes pages fast. Google's published position is that it judges content by whether it is accurate, useful and adds something, however it was made; what its spam policy names is scaled content produced mainly to rank. Mass-generate pages nobody would sign and that is what you are producing.

Marketing guys will offer to generate hundreds of blog posts for you. Pages like that can show impressions for a while and then get filtered out under the scaled-content policy, and the traffic goes with them.

Write pages a person would sign: one real service per page, in your words, checked by you. Publish at the pace you can actually review; there is no magic number of posts a week and no rule about the time of day.

Google's engineers are way smarter than any one of us. Play by the rules of the platform. It's the same principle as the storefront: build it the way Google wants to find it.

The third place: being found in AI search.

People are now asking ChatGPT "best structural engineering firm near me". It's a small share of searches so far. At one client firm, 5 of the 26 web leads that Google Analytics tracked in August 2026 came from ChatGPT, so it is worth being in those answers.

Google's own guidance for its AI features says no special optimization is needed and inclusion is not promised; the other assistants publish less, and what they reward shifts.

A lesson about tricks: every year a shortcut appears that works for a few months and then stops, and the months spent on it are gone. Tactics without durable evidence are not a plan; the pages about what you actually do keep working.

What lasts is what Google rewards too: genuinely useful pages about what you actually do. Build useful pages; neither search rankings nor inclusion in an AI answer is assured.

Pay for traffic once the storefront and the instruments exist.

Never before. Google Ads for the searches happening now, Meta for the relationships, outreach if the budget is tight.

Two ways to get traffic, and the three levels of paying for it.

Organic: being found for the exact service keyword your prospect is typing. Everything in the last section. It compounds, and it's slow to start.

Paid: if you're not being found organically yet, you can pay your way into that placement today, for the exact same searches.

Google and Meta make the majority of their revenue from advertisers. Their platforms are built to take your money and give you the placement. Google Ads is the easiest way in.

Three levels, in order. Beginner: Google Ads. Medium to pro: Meta. The budget version: outreach. Do them in that order, and none of them before the storefront and the tracking exist.

Google Ads: know your numbers before you spend a dollar.

The example below is hypothetical: round numbers chosen to show the arithmetic, not a forecast. Your market sets your numbers; do the math in your own units.

Google Ads runs as an auction, so what a lead costs depends on who else is bidding in your market. I rarely see engineering firms advertise here, which is why the auctions I've run have been inexpensive. Check your own market before you assume the same.

The math is yours to do: what does a site assessment pay you? What does a project pay? What's your close rate on the phone?

If a paid lead costs less than what it earns you on that arithmetic, the ads pay for themselves. If it costs more, they don't, and the tracking is how you find out which.

When the arithmetic holds, you may not be maximizing profit, but you are bringing in work you otherwise wouldn't have.

Cost per lead: Say $60. For scale, one measured figure: one client firm paid $52.64 per lead in ad spend, as its Google Ads account counted them, in July and August 2026

Leads that book a paid assessment: Say one in four. Yours depends on how the phone is answered and what you charge

What one assessment pays you: Say $500. Use your own fee

The result: Four leads cost $240 and bring one $500 assessment, before your time and before any project that follows. Change any line and the answer changes

Fill in the right column with your own numbers before the training. Keep the pipeline full first; maximize margin second.

Be honest about the quality, and about why the market is open right now.

In the accounts I've run, search has mostly brought urgent, lower-ticket homeowner work, like the load-bearing wall assessment. Useful for keeping the calendar busy; not usually the relationship that carries a firm for years. What search brings you depends on the searches you target.

So my recommendation: Google Ads for the now-intent work. The relationships that bring project after project come from somewhere else, next slide.

Why I think the opening is there. This is my read, not a measured fact: I see more small engineering firms than I used to, and few of them advertise on search.

Fewer bidders usually means cheaper placements, if you are the one willing to spend and measure.

Meta: interrupting the B2B relationships that bring many projects, not one.

A long-term engineering partnership can start through more than one route, including search, referrals and direct introductions. On Facebook and Instagram you interrupt someone's day to let them know you exist. That's pattern interruption, and it's how you reach people who aren't searching.

The targets: architects, general contractors, developers. One relationship there can bring in many projects over the life of the partnership, worth years of work.

An ad I'd run to architects: "Is your structural engineer about to retire? Are you looking for a new one?" Then the offer: a free assessment showing how few comments we get back from the city, and how quickly we respond to the ones we do.

It's harder than Google and there's more room to scale. Do it after search is working and the tracking can tell you whether it paid.

Outreach: the budget version. One relationship a month changes a year.

If you're tight on budget, build the list. Every architect, general contractor, interior designer and kitchen designer in your state. Whoever your B2B clients actually are.

Write to them. A real note about a real thing, not a blast, even when the volume is high. A handful a day, done properly, is enough.

Not everyone reads email, and that's fine. You only need one or two replies a week. Because these are B2B relationships, one new one a month can matter more than a month of small jobs.

Keep the list in the CRM, not a spreadsheet, so every reply lands in the pipeline and gets a follow-up. That's section five.

Every lead those four things produce has to land somewhere.

Not your inbox. Not a sticky note. A marketing CRM: the pipeline, every conversation, and a gate on the phone line.

A marketing CRM, which is not the software you run projects in.

Collecting the data is one thing. Having a pipeline for every marketing initiative is what lets you do sales well: track how many people are coming in, and organize them by stage.

At a glance you know what stage each prospect is in. Do you have to follow up with this person? Do you not? Nothing falls through because it lived in someone's head.

Before recording calls, establish the notice and consent your firm needs and set up the line accordingly. The worst thing is someone calls you, you forget the conversation, and you don't know how to follow up. It looks extremely poor on you as a professional. With the recording you go back and follow up exactly.

A marketing CRM is built to move a stranger to a signed client. Your project software is built to run the job after they sign. Different tool, different job.

The pipeline makes your sales process visible, and you can only fix what you can see.

Every inquiry is a card, and you drag it across the stages as it moves. New, Contacted, Site visit, Proposal, Won. Use your own stage names; the point is that every lead is somewhere.

When it's in front of you like this, you see where deals die. Is it the first call? The proposal? That's your sales process made visible.

Every conversation in one place. The website form, email, texts, the phone call, one thread per person, so your team answers from one screen instead of four apps.

The CRM talks back to the ads. When a lead becomes a real conversation or a signed project, that can be sent back to Google, so the ads learn what a good lead looks like. Found, captured, answered, counted, improved: one loop.

Protect the front desk: the gate on the line.

In its worst week, one client's tracked phone line took 99 calls, three in four of them from callers the system could not identify, while a setting on the line was hanging up on real callers. The call log is what showed the size of it.

We put a gate on the line: press one for a new inquiry. A gate is a filter, not proof of a person: automated callers can also send a keypad response. In the same call log, calls fell from 99 that week to about 18 a week, and unidentified callers from 75 to about 6.

The gate reduced unwanted calls in this account. The inquiry count uses a separate rule: a caller new to the CRM whose call lasted at least a minute, or someone who submitted a form or chat, with duplicates removed. The rule does not establish that every inquiry is a new customer or a suitable project.

The lesson for you: track the line first. You can't fix a phone problem you can't see.

What "built" looks like. Score your firm on all five, honestly.

01 Storefront: As good as the sites from your own search. Design, a fluid journey to a real form, the SEO structure. Live, not perfect.

02 Instruments: Search Console, Analytics and Tag Manager set up and wired. You know last month's impressions.

03 Being found: A profile that represents the location or service area you actually serve, the right category, reviews asked for after every project, a page for every service, written like a person.

04 Traffic: Google Ads for now-intent with your own math done. A B2B list, and a way to reach it on Meta or by email.

05 The system: A marketing CRM with a pipeline, every conversation in one place, calls recorded, a gate on the line.

The lowest score is where your next thirty days go. If everything is a 1, start at the top: the order matters.

The order of operations. Each step makes the next one pay.

Search your city. Write down the three map-pack firms and the top organic sites for "structural engineer near me" and "load-bearing wall removal near me". That's your bar.

Fix the storefront. Design, the journey, the SEO structure, and a form with the five to seven questions you ask on every first call. Ship the B-plus in two weeks.

Set up the three accounts the same week the site goes live. Search Console, Analytics, Tag Manager. Look at them monthly.

Fix the profile. Right category, the location or service area you actually serve (a displayed office only if it meets Google's requirements), video verification. Then ask every past client for a review, and every new one at close.

Write service pages like a person, at the pace you can review them. One page per real search. Don't mass-post pages nobody would sign.

Only then, pay for traffic and land it in a CRM. Google Ads for now-intent with your math done; the B2B list for Meta and outreach; a pipeline so nothing falls through.

The easy part is marketing. The hard part is your job.

Getting you the leads is the easy part. That's what I do; it's where my expertise lies.

The hard part is fulfilling good service to those leads. Doing it badly is the fastest way for a market to learn you're not a reputable provider. No amount of marketing fixes that.

If you have the time to build it yourself, start here. Around $50K a month is a guideline for considering managed help, not a cutoff. Storefront. Instruments. Found. Traffic. The system.

If you want help, work out which of the five you do not have time to build or manage. That's the part I do, and I take on one firm per primary engineering discipline in a metro. If there is an existing conflict, I'll explain it; we can still have a useful conversation.

The data doesn't lie. Go look at yours.