Greg Marshall
You have decided GoHighLevel is the right tool. Maybe you have already signed up. Someone has offered you a snapshot for your trade, or you have watched enough videos to be fairly sure you could set it up yourself over a couple of weekends. Either way the pull is the same: get something live, start using it, and sort the details out as you go.
That is how most owners get into GHL, and it is why most of the accounts I go through are quietly underperforming a year later. The software is rarely the problem. The decisions that make a system work were skipped in the rush to get going, and ever since then the system has been shaping the business rather than the business shaping the system.
This article walks through the seven decisions I would make before touching the software. In the video I go through them with a worksheet open, but you do not need it to follow along.
The two ways in, and what each one costs
There are two ways a service business usually ends up in GoHighLevel.
The first is a snapshot. A pre-built account for your niche, often free or cheap, and you can be up and running in a few hours. The appeal is obvious. The catch is that it was not built for your business, your process or your team, so you end up fitting the business to the snapshot rather than the other way round. Snapshots also come with a lot of bloat: workflows, calendars and custom fields for things you simply do not do. In my experience about half of what arrives in a snapshot never gets used, and the half you keep has to be bent to fit how you actually sell.
The real price of a snapshot is not the licence. It is that a system that fits badly converts fewer of the enquiries you are already paying for, and that bill arrives every month for as long as the system is running.
The second way in is a blank slate. You take an empty account and build the system around your process. It fits, there is no dead weight, and the act of building it forces you to think through how you want things to work. The cost is that it is slower, it needs a proper understanding of your process, and it takes either money or time.
If you do it yourself, the honest estimate is 30 to 40 hours to learn the system and build something usable. Value your time at, say, £50 an hour and that is £2,000 of your time before a single lead has gone through it, and the result is usually something you would grade a C. There is also a quieter cost I think of as paralysis: you do not know what the software can do, so you cannot tell whether the thing you want is possible, and the limits of your own imagination become the limits of the system.
I am aware that a GHL builder telling you to build custom sounds like a pitch. That is not the point of this article. Whichever route you take, the seven decisions below need making first, and if you make them properly the snapshot question largely answers itself.
Seven things to work out before you touch the software
1. Snapshot or blank slate (and why to leave it until last)
This is the decision everyone makes first, usually in about ten minutes, usually on price. I would leave it until the end. Once you have done the other six, you will know whether your process is close enough to standard for a snapshot to fit, or whether you would spend the next year fighting it.
2. Find out where your leads actually die
Take the last 90 days of enquiries. Remove everyone who went ahead. For each of the rest, work out the point at which the sale broke down. Did they enquire, get a reply, reply again, and then go quiet? Did you send a quote and never hear back? Did they book a visit and not turn up?
In most businesses only about one in five people come back with a clear yes or no. The rest dwindle away, and nobody in the business can say at which stage. Doing this exercise tells you where you are weakest, which is also where the biggest improvement is sitting. Everything you build afterwards should be aimed at that stage first. In most service businesses it turns out to be after the first fortnight, with no-shows and quotes that went quiet, and I have written separately about what to send at that point.
3. List every way an enquiry comes in
Website forms, ads, social media, the phone, email, WhatsApp, walk-ins, referrals that arrive as a text to your personal mobile. Write down every single one.
You do this because each route has to be plugged into the system and given a path. An enquiry that lands by email needs to end up in the same pipeline and the same follow-up sequence as one that came through the website form. If a route is missing from the list, the enquiries that come through it will be handled by hand, or not at all, for as long as the system exists.
4. Write down what really happens now
Not what should happen. What actually happens, stage by stage, including the manual bits. Who is accountable for each step? How long does each step take? You might find you already have a different CRM with a few automated messages, but nothing much behind it.
A line like “half our leads come in by email and whoever is free messages them back when they get a minute” is not an embarrassment. It is a finding. It tells you exactly what the first automation should do: when an email enquiry arrives, it goes into the system and the first message goes out without anyone having to be free.
5. Design the ideal version, including the unhappy paths
Now think about how everything would work in an ideal world, step by step. The mistake most people make here is only planning the happy path: enquiry comes in, they reply, they book, you meet them, you quote, they say yes.
Real pipelines are mostly unhappy paths. What happens if they do not reply to the first message? How many follow-ups go out, and over how many days? If they book a visit and do not turn up, what happens next? If the quote goes quiet, who chases it and when? Each of those needs a planned route, because the leads that take them are the ones you found dying in step two.
6. Decide what dies and who owns what is left
A new system is a chance to retire things. The spreadsheet nobody updates. The email threads about leads. The separate CRM. The form tool you pay for separately. GoHighLevel has email, phone and social messaging in one place, and it will build your forms, so a lot of the surrounding software can go.
For everything that stays, put a name next to it. Not a team, a person. The step with no owner is the step that leaks.
7. Cost it honestly, in money and in hours
How much will this cost, and how many hours will pass before it is actually live and being used? Be honest about both, and compare them with the cost of getting it wrong. A system that took 40 hours to half-build and then converts poorly is the most expensive option of all, because you pay in time up front and in lost enquiries every month afterwards.
What the wrong choice actually costs
Most owners rush because they want something to get started on and plan to figure it out from there. The trouble is that these systems take a lot of time and energy to get running, so once you have installed a snapshot or started building, you have already committed before thinking it through.
Here is the arithmetic that matters. Say a system that fits badly converts five of your enquiries a month, and a system built around how you actually sell converts nine or ten of the same enquiries. If an average job is worth £2,000, that is £8,000 to £10,000 a month you are not seeing. It does not show up anywhere as a loss. It just looks like a quiet month, and it quietly accrues for as long as the system runs. That gap is why a properly thought-through build more than pays for itself.
What to do this week
Do step two. Pull the last 90 days of enquiries out of wherever they live, mark each one won or lost, and for the lost ones add a column for the stage where it stopped. It takes an afternoon and most owners are surprised by what they find.
Then do step three and list every way an enquiry comes in. Those two lists will tell you more about what to build than any demo or snapshot tour, and they are what I would want to see before giving anyone a recommendation. When you are ready to write the whole process down, this is how I would plan the system on paper before anyone builds it.
When not to bother
If you are one person with a handful of enquiries a week, a snapshot or a tidy spreadsheet is fine. The difference between five and nine conversions a month does not apply when the volume is not there, and your time is better spent on the work.
If you already have a snapshot installed and it is converting well, do not rip it out on principle. Find out where it leaks first. You can do a fair amount of that yourself by exporting the workflows and running them through an AI audit. Snapshots can work. The problem is only ever that you adapted to the system without noticing what the adaptation cost.
If you already have GoHighLevel in place, whether that is a snapshot someone sold you or a build you started and never quite finished, the quickest way to do steps two and four is to have someone go through it. That is what the free audit is: I go through your workflows, calendars, forms and pipelines, and you keep the report whether or not you do anything with it. Find out where yours is leaking.