It is your fourth discovery call this week and you are explaining what the product does in slightly different words than you did yesterday. Three demo requests from last Tuesday are still sitting in a shared inbox. The launch you planned for September is now a launch you are planning for November.
There is no revenue number or funding round that tells you when a startup should hire a marketing agency. There are signals that founder-led marketing has stopped scaling, and they are much more reliable than a milestone.
Founder-led marketing is the right default, until it is not
Early on, the founder is usually the best marketer the company has. You know the customer, you can explain the product without notes, and every conversation teaches you something. Handing that off too early is a mistake: an agency hired before you know who buys is being paid to guess, and the guesses will be expensive.
So the question is not whether you should be doing your own marketing. You probably should have been. The question is whether doing it yourself has started costing more than it returns.
Five signals it has stopped scaling
You are the only source of pipeline. Trace your last five deals back to where they started. If every one of them runs through your own network, your own posts or your own calls, the company's growth is capped at your calendar.
Inbound arrives and nobody follows it up. Not because nobody cares, but because nobody owns it. A demo request answered a week late is mostly a lost demo request.
You rewrite the same explanation for every call. There is no positioning line written down, so it lives in your head and comes out slightly differently each time.
A launch slipped because nobody owned it. The product was ready. The page, the email and the docs were not.
You cannot say which channel produced your last five customers. Not as a percentage, just as a list. If you cannot, you are deciding where to spend time on instinct.
One of these is a busy month. Three or more, for more than a quarter, is a pattern.
What to fix yourself first
Some of this is not agency work at all, and you will get more from any outside help if these are done first:
Write the positioning down. One paragraph: who it is for, what it replaces, why it is better for them. Use it on every call, the homepage and the next deck. This costs an afternoon.
Set a follow-up rule and give it an owner. Every inbound request answered within one working day, by a named person, logged in a CRM or even a shared spreadsheet.
Record where every deal came from. One column, filled in when the deal opens.
Pick one channel and do it properly for a quarter, rather than five channels badly.
A note on following up by email: under Canada's Anti-Spam Legislation (CASL), a commercial message needs consent, clear sender identification and a working unsubscribe, and the onus of proving consent sits with the sender (sourced from the CRTC). An inquiry from a prospect can give you implied consent, but only for a limited period the Act sets out, so log when and how each lead reached you. This is marketing guidance, not legal advice.
What genuinely needs outside help
Outside help earns its fee on work that needs sustained hours or a skill nobody on the team has:
A content engine that publishes every month whether or not the founder had a quiet week.
Technical SEO on a site that engineering built for the product, not for search.
Paid acquisition that needs managing weekly and judging on pipeline rather than clicks.
Lifecycle email and CRM flows that nobody has had time to build.
Launches with an owner who is not also shipping the feature.
What it cannot do is work out who your customer is for you. That stays with the founder, and any agency that offers to discover it from scratch in month one is selling you its learning curve.
Agency, fractional or a first hire
Once you know the work is real, the next question is what shape of help fits it. Broadly: a first hire makes sense when the work is a full-time job and you can afford to train someone; a fractional lead or an agency makes sense when the work is real but not full-time yet, or when you need it running before you can hire. What each option actually costs in Canada breaks the three down with the numbers.
How to start without betting the runway
If you are not sure yet, do not sign anything long. Start with something fixed in scope that tells you whether outside help is worth it. My version is the Diagnostic: $1,500, one time, for read access to your analytics, Search Console, CRM and site, a written diagnosis, a sequenced 90-day plan and one 60-minute call. If you then start a retainer within 30 days, the whole $1,500 comes off month one. What the Diagnostic covers is published, and the retainers after it run month to month, with no annual contracts and two months' notice. For pre-seed founders who need marketing built before there is revenue to pay for it, there is also a Founding Marketer arrangement, paid in equity, with a six-month minimum and limited places.
Whichever way you go, know what you will hand over on day one. What to hand over to a marketing agency is the checklist, and doing it before you sign makes the first month count.
The short version
Fewer than three of the five signals: keep doing it yourself, and fix the basics above.
Three or more, for a quarter or longer: the work is real. Decide what shape of help fits it.
Not sure: buy something small and fixed before anything monthly.
So, of those five signals, how many did you recognise?