A logo is not a brand, and starting with one is the most common and most expensive mistake a new business makes. Here is the order that actually works.
Most startups approach branding as a design purchase. They commission a logo, pick colours they like, and discover a year later that nothing hangs together — the website, the deck, the packaging and the social presence each look like they belong to a different company.
The problem is sequence. A logo is a conclusion, not a starting point. Everything visual should be answering questions that were settled beforehand, and when those questions have not been asked, the design has nothing to be accountable to except taste.
Settle the strategy before opening a design tool
Four things need answers, and they do not require a designer. What do you actually do, described so a stranger understands it? Who is it for, specifically enough to exclude people? Why would someone choose you rather than the obvious alternative? And how should the business sound — plain and direct, warm and personal, precise and technical?
These answers become the brief. They are also what makes design review productive: instead of arguing about whether a mark is nice, you can ask whether it suits a business with this positioning talking to this audience. That is a question people can actually answer.
Positioning means choosing what you are not
The hardest part of this work is exclusion. A brand that tries to appeal to everyone gives a designer nothing to work with, and gives a customer no reason to remember it.
Being the careful, slower, more thorough option is a position. Being the fastest and most affordable is a position. Being adequate at both is not, and it produces the beige identity that most new companies end up with — safe, forgettable and indistinguishable from three competitors.
What a visual identity actually contains
A single logo file will not carry a business. Within the first month you will need the mark in contexts it was never drawn for: a square social avatar, a wide website header, a single colour on an invoice, an embroidered shirt, a browser tab sixteen pixels across.
- Primary mark, plus stacked and horizontal lockups for different spaces.
- A monochrome version that works in one ink, and a reversed version for dark backgrounds.
- A colour palette with defined roles — not just a set of swatches, but which colour is for what, and which combinations meet accessibility contrast requirements.
- Typography: a heading face, a body face, and rules for sizes and weights that someone non-technical can follow.
- Guidance on imagery and iconography, so photography and graphics do not drift apart over time.
- Clear space and minimum size rules, which is what stops the mark being crushed into a corner.
Testing at the extremes is where identities fail. We check marks at favicon size, in one colour, on dark backgrounds and against contrast requirements before guidelines are finalised, because discovering a problem afterwards means redrawing rather than adjusting.
Write the guidelines so someone will read them
A brand guideline document nobody opens has no effect. The ones that get used are short, practical and written for the people who will apply them — an office manager formatting a proposal, a contractor building a landing page, a supplier printing a sign.
That means plain language over design vocabulary, examples of correct and incorrect use, and files organised so the right one is easy to find. A ten-page document people follow beats a sixty-page one they ignore.
Consistency is what turns marks into recognition
Recognition is built by repetition. The same mark, the same colours and the same voice appearing consistently is what eventually makes a business familiar. Changing them because the team has grown bored resets that process, usually at the exact point it was starting to work.
Internal fatigue is not a reason to rebrand. You see your own branding far more often than any customer does. By the time it feels repetitive to you, most of your market is only beginning to recognise it.
Rolling it out across touchpoints
An identity only exists where people meet it. That means the website, email signatures, proposals and invoices, social profiles, packaging, vehicle livery, signage and any advertising — and they should change together rather than one at a time.
A staggered rollout is where consistency dies. The website gets the new mark, the proposal template keeps the old one because nobody owns it, and eighteen months later there are two versions circulating. Before launching, list every surface that carries the brand and assign each one an owner. The list is always longer than expected, and that is the point of writing it down.
What this costs a startup
Branding is sold across an enormous price range, and the range is confusing because the deliverables are described in similar language. A cheap logo package and a full identity programme are not the same purchase, even when both are called branding.
What a young business genuinely needs is narrower than an agency rate card suggests: the strategy questions answered, a mark that works at the sizes you will actually use, a defined palette and type pairing, and a short guideline document. Sub-brands, motion systems and extensive photography direction can wait until there is a business to apply them to.
What this looks like as a sequence
- Strategy — what you do, who for, why you, how you sound.
- Naming and verbal identity, if the name is not already settled.
- Visual identity — mark, lockups, palette, type, tested at real sizes.
- Guidelines — short, practical, with files organised for the people using them.
- Application — website, documents, signage, social, rolled out together rather than piecemeal.
Skipping straight to step three is what produces the identity that needs replacing eighteen months later. The earlier steps are cheaper, faster and considerably less fun than looking at logo concepts, which is precisely why they get missed.
Need help with this?
We deliver this work for businesses across Canada every week.
Talk to us about your project
