Three days before JB Marketing's first anniversary, our co-founder Ben sat down with Revline's Check-in Podcast. The conversation covers how the agency went from two freelancers to a team of six in a year, where the clients actually came from, and the internal portal that started life as a list in a spreadsheet.
The full episode is below. If you'd rather read, the main points are written up underneath it.
If you take one thing from it: build the smallest version of a system you need today, then let the business tell you what to build next. A portal with no clients in it is a hobby.
From two freelance books to one agency
Ben has worked in Google Ads since an apprenticeship straight out of college, moving from a large agency to a smaller performance-focused one and staying in the platform ever since. Like most people in marketing, he had freelance clients on the side of the day job. So did James.
JB Marketing is what happened when those two freelance books merged. The name is literally James and Ben. There was no cliff-edge moment of quitting with nothing lined up. The freelancing ran alongside employment until the combined business was at a level where switching over full time was comfortable.
"It wasn't job, then no job, then JB. It was job and freelance, then the freelancing merged with James's and became JB. There wasn't a drop."
Ben, on the transitionThe freelancing mattered for another reason. A Google Ads specialist in a day job never touches client acquisition or pricing. Freelancing was the training ground for the business side before there was a business.
The first real step from freelancing to agency was standardised pricing. As a freelancer Ben had been taking random projects at random retainer amounts. Having an offer, with a price structure attached to it, is what made it a company. Some clients left at that point. That was fine.
Referrals are still the number one channel
The earliest clients came from partnerships with ecommerce and business communities. JB became the go-to Google Ads experts inside a few of those groups, answering questions, running ask-me-anything sessions and course-style content. People met the agency through a source they already trusted, and that converted.
A year in, referrals from existing clients are still the main source of new business. One client has referred siblings and parents, each running their own brand. Ben's view is that a client trusting you enough to send their family is a better performance metric than anything in a dashboard.
How to get more of them
Being good at the work comes first and nothing replaces it. After that, it's about timing the ask, and that's where the operations hire two months ago changed things.
- Customer satisfaction scores are collected quarterly, covering communication and performance.
- A red, amber, green rating sits on every client. It's set jointly by operations and the account manager, and it reflects both performance and the relationship. Good numbers with a strained relationship still lands you at amber.
- Only green clients get the ask. A client who has just rated you highly is the one to ask for a referral or a review. An amber or red client is not, however good the ROAS looks.
- The deal is sweetened. A successful referral takes $1,000 off the referring client's next invoice, or gets paid out directly. For some contracts that's half the month's fee or more.
Overcommunicate until you know what they want
At the start there was no communication process at all. Clients got answers when they asked questions. The lesson since then is that overcommunication beats undercommunication every time.
The baseline now is a Monday, Wednesday and Friday update on every account at minimum, covering what's being worked on, how performance looks and anything needed from the client. On top of that, the team deliberately starts conversations about promotions, new products, changes in direction and what the client's other agencies are doing, because a Google Ads account that doesn't know about next week's sale is going to have a bad week.
The other half is finding out how each client wants to be talked to. One client runs entirely on WhatsApp, so they get shorter messages more often. Some want a weekly or fortnightly call. Some want no calls at all, just a weekly note and a monthly summary.
"If we don't know what they like, we'd prefer to overcommunicate. If we know what they like, we tailor how we talk to them and how we work to what they want."
Ben, on client communicationStart on Google Sheets, build the system later
This is the part of the conversation the episode is named after, and it's the most useful bit for anyone running a service business.
For the first five or six months, JB had no clear view of its own profit. Nobody was tracking what it cost to fulfil a client. There wasn't a reliable top-line revenue figure or a count of accounts being serviced. During the months where growth was strongest, that got uncomfortable, because there was no central place any of this data lived.
Version one of every tool the agency now uses was a Google Sheet or a Google Doc. The SOPs were Docs. Finance tracking was a Sheet. The outreach CRM for Instagram and LinkedIn still is a Sheet, because it does the job.
How the portal grew
What the team now calls the portal is an internal website that runs the company. It didn't start that way. Each version was a response to a problem the business had at the time.
- Version 1A list of clients, with their billing details and retainer amounts, in one place.
- NextA revenue dashboard on top of that client list, showing churn, margin, growth and profit, because the numbers were needed and the client data was already there.
- With the teamA team page for staffing costs and time-off requests, once there were people to staff and time off to request.
- TodayAround 250 versions later, it holds per-client and per-employee P&L, tooling costs down to the API key, every SOP, audit builders, keyword research, and a connection that gives each team member's AI assistant live access to all of it. It integrates with Slack, Granola call notes, Google Calendar and Gmail. It even hosts advertorials and landing pages for client campaigns.
The practical result is that Ben can ask what's been done on any account this week and get an answer pulled from what the portal knows was actually actioned. New hires get one connection and immediately have Google Ads access, research tools, audit builders, SOPs and client context. The feedback from recent hires is that it speeds up the day-to-day of running Google Ads a lot.
Why not buy the tools?
There are tools that do the Google Ads piece. None of them does the finance piece, the HR piece and the note-taking piece in the same place, and separate tools don't talk to each other. The cost creeps too. One Google Ads tool Ben looked at was around £70 a month per user, which at four users is £280 a month for a single function. The whole portal runs for under £150 a month and covers everything, branded and built around how the team actually works.
The trap on the other side
Ben is candid that building is a rabbit hole. It's easy to keep adding things that are nice to have. His rule now is that something only gets built if it's a request from the team, a bug the team has hit, or a change with a measurable impact on the business. Anything else waits.
He's also clear that the order matters. With the knowledge he has now, he could spend hundreds of hours building the portal before landing a single client, and it would be worthless, because there'd be nothing to put in it. The portal only made sense once there was a team to use it. At that point the maths flipped: it now saves something like ten hours per team member per week, which outweighs every hour that went into building it.
"You've got to find the minimum viable product. Whatever your eventual product is, you need the smallest, easiest implementable version of it first."
Ben, on when to buildCustom or off the shelf?
Asked whether a new service business should lean custom or buy something established, Ben's answer is that it depends on your knowledge. He'd built websites before and understood the logic of how software fits together, which made building viable. For most businesses, Sheets cover enough for a long time. The moment you need integrations between things is the moment you need real tooling, and at that point it's sometimes easier to build a small connector than to find a niche product that happens to join the two tools you use.
Two pieces of advice
If you're just starting
Client acquisition is a volume game. If you message ten people and get two decent conversations, message a hundred and get twenty. The best salespeople still have a lot of conversations for every close, so ten unanswered messages is not evidence you're bad at this.
Ben also pushes back on the idea that going independent is the risky option. You can lose a job with one mistake. With five, ten or twenty clients, you'd have to lose all of them to hit zero. Spread across enough clients, it's the more secure position, not the less.
If you're trying to scale
Track the numbers and let them make the decisions. JB knows exactly how many clients it needs before it starts hiring again. Early on, decisions like that were made on gut feel. With capacity and margin tracked, even in a spreadsheet, you know when to pull the trigger rather than guessing.
Watch the full episode
The whole conversation is 28 minutes and worth it if you run or are thinking of starting a service business. You can watch it on YouTube, and you'll find more episodes on Revline's channel. Thanks to the Revline team for having Ben on.
If you'd like to know more about who you'd be working with at JB, or want to see what the communication and reporting rhythm looks like on a real account, the discovery call is the easiest place to start.
Quotes are lightly edited for clarity.