Custom Logo
Design Agency

A logo design agency draws an original mark, proves it works everywhere it will actually appear, and hands over every file you will ever need. Those three things are what separate professional logo design from a picture somebody sent you, and a large share of the marks we are asked to fix fail at the second or the third.

What the work actually involves. Research into the category and the marks already occupying it, so yours stands apart instead of blending in. Broad sketch exploration before anything gets refined. A small number of strong directions presented with the reasoning attached. Then testing at every size the mark will face, down to a favicon, before it is shown to anybody. Then the full file set and the guidelines that keep it consistent.

Why clients choose us for it.

Every mark is drawn from scratch. No templates, no stock symbols, no generated assembly. That takes longer and it is the only way to end up with something you can actually own.

Concepts are tested small before they are presented. A mark that only works at presentation scale is a mark that will fail on a phone, an app icon, and an embroidered shirt. We check that before you see it, not after you approve it.

You own all of it outright. Every file, every format, with the rights transferred. Covered properly further down because it is the question buyers most often forget to ask.

Senior people do the drawing. More than 100 original logos delivered across a decade of design work, and the person sketching is the person presenting. You can see the marks themselves and the people behind them before deciding anything.

The one thing worth saying up front. A logo does not need to explain what the business does. It needs to become the thing people recognize, which is a harder job and a different one, and confusing the two is the most common reason a logo project goes in circles.

Five logo design services, covering a mark's whole life from the first sketch to the guidelines that protect it. Most engagements use three or four of them.

Custom logo design. An original mark drawn for the business it represents. Category research, broad exploration, a small number of resolved directions with their reasoning, testing at favicon size, and delivery of every file owned outright. The exploration is deliberately wide before it narrows, because the first credible direction is rarely the best one available.

Logo redesign. Modernizing a mark without discarding the recognition already built into it. Detail simplified, proportions corrected, and the files rebuilt properly so it holds from a favicon to a storefront. Where the change needed goes past the mark itself, it becomes a rebranding conversation and we will say so.

Wordmark logo design. Custom letterform work for businesses whose name should be the mark. Drawn from the letters up instead of from a typeface down, so the name cannot be reproduced by typing it. Where the name itself is still unsettled, naming comes first.

Brand marks and symbols. For the places a full logo will not fit, meaning app icons, avatars, favicons, and small physical applications. Delivered as part of a complete lockup system alongside the wordmark it pairs with. A symbol drawn for sixteen pixels is a different drawing from the same symbol at two hundred, and both get made.

Logo guidelines. Clear space, minimum sizes, palette values for print and screen, approved lockups, and the misuses that damage a mark fastest. Short enough to be read and specific enough to settle arguments. Written so a printer or a new hire can follow them without a designer sitting beside them.

Most projects use a combination instead of one item. A young business with a settled name usually takes custom design plus guidelines. An established business with recognition to protect usually takes the redesign and nothing else. A business whose name is the whole story takes the letterform work and a symbol for the small spaces. The mix gets decided after we look at where the mark actually has to appear, and not from a menu.

Where a logo is one part of something larger, meaning a full identity system with typography, palette, imagery direction, and layout rules, or the collateral and templates a business runs on daily, that is scoped as brand work instead. A logo on its own solves less than most people expect, and we would rather tell you that before you buy one.

Distinctive, simple, scalable, appropriate, and ownable, in that order of difficulty. Everything else people argue about in a logo review is preference dressed as principle.

What each of those actually means.

Distinctive. Different enough from the marks already in your category that somebody could pick it out of a lineup. Most logo briefs ask for the opposite without realizing it, since the reference images clients bring are almost always competitors. A mark that looks like the category is a mark nobody remembers.

Simple. Reducible to its essential shape and still recognizable. The test is whether you could describe it over the phone, or draw it badly from memory and still have it read. Detail is what fails first at small sizes, and a mark carrying more than a handful of distinct shapes is usually carrying decoration somebody was reluctant to remove.

Scalable. Works at sixteen pixels and on a building. Sixteen pixels is roughly the height of a lowercase letter in a line of body copy, which is the honest scale a favicon has to survive. That means no hairline strokes, no gradients that muddy in one colour, no tight counters that fill in, and no drop shadows.

Appropriate. Belonging to its industry without disappearing into it. What reads as premium in one category reads as wrong in another, so we establish the conventions across the industry and then break them deliberately and selectively.

Ownable. Original enough to defend, which is a legal question as much as a creative one and gets its own answer further down.

What a good logo does not have to do. Explain the business, illustrate the product, contain the industry's obvious symbol, or mean something on day one. Symbols start with no meaning and earn it through repetition. The most recognizable marks in the world are arbitrary shapes that became meaningful because people saw them ten thousand times.

Which is why we do not judge a mark on whether it communicates the strategy. We judge it on whether it is distinctive, holds up small, and can carry meaning once it has been used for a year. The strategy belongs in the positioning and messaging, and the category research is what tells us where the space actually is.

Six shapes, and the right one depends on your name, your category, and where the mark will be seen most. Most businesses end up with a combination, and the decision is more practical than aesthetic.

Wordmark. The name itself, drawn as the logo. Right when the name is distinctive, reasonably short, and worth building recognition around directly. It is the most efficient choice for a young business, because every use builds name recognition and mark recognition at once. Custom letterform work matters here, since a name set in a stock typeface can be reproduced by anybody with the same font, and where the name itself is still in play that decision comes first.

Lettermark or monogram. Initials, used when the full name is long or awkward. Works well in small applications and needs care, because two or three letters in a box is the most crowded space in logo design.

Pictorial mark. A recognizable object, simplified. Immediate to understand and risky, because a literal symbol can date quickly or lock a business into what it did at the start.

Abstract mark. A geometric or organic form with no literal reference. Maximum flexibility and maximum patience required, since it means nothing until repetition gives it meaning. The right choice for businesses that will outgrow a literal description.

Emblem. Type contained inside a shape, like a crest or a badge. Strong presence and traditional authority, and the weakest at small sizes, which makes it a harder choice than it used to be.

Combination mark. A symbol and a wordmark designed to work together and apart. This is what most businesses actually need, delivered as a lockup system with horizontal and stacked arrangements, plus a standalone symbol for the places a full lockup will not fit. It is also the shape that extends most easily into everything that comes after it, which is why it appears in most of our brand engagements.

How we decide. Where the mark will appear most, whether it needs to work without the name beside it, how long the name is, what the category already looks like, and whether you have the patience and the exposure for an abstract mark to earn its meaning. Then it gets built as a system instead of a single file, since a lockup that works in one arrangement and fails in another is half a job.

A focused logo project runs three to five weeks. A logo as part of a full identity system runs six to ten. Logo design cost scales with the same variables as the timeline, so the two are worth answering together. Cost scales with how much exploration and how many applications are in scope, and a logo on its own is the smallest branding engagement we take.

What drives the timeline and the number.

• Whether strategy is settled. If nobody can state what the business is for and who it serves, that gets resolved first. Designing a mark against an unfinished position is how a project loops through batches that feel like taste and are actually strategy.
• How much exploration the category needs. A crowded market takes more sketching to find a space nobody occupies.
• Whether the name is final. A wordmark cannot be drawn against a name that might change.
• The application set. A mark, a lockup system, and a favicon is one scope. Add signage, vehicle livery, packaging, embroidery, and an app icon and it grows.
• Whether guidelines are included, and how thorough they need to be.
• Approval structure on your side. The single largest cause of a slipped timeline. One decision-maker moves twice as fast as a committee, and it is worth being honest about at kickoff.
• Trademark screening, which has its own timeline nobody controls.

What we will tell you plainly. A logo on its own frequently is not what a business needs, and buying one when the actual problem is positioning means paying twice. Where that is the situation, a scoped consultation is a cheaper way to find out than a design project. And if you are working to a launch date, say so on the first call and you will get a straight read on whether that date and that scope can coexist.

The other honest note is about the bottom of the market. There are marketplaces and generators producing a mark for very little, and they are covered properly further down, because the comparison is not really about price.

Ideas come in batches, we discuss them together, and we keep going until the mark is right. There is no revision limit.

Why there is no cap. The moment a contract says three rounds, the conversation changes. A client starts rationing feedback, saving a round in case something worse comes up later, and quietly not mentioning the thing that actually bothers them. That is how a mark gets approved that nobody in the room really likes. Removing the ceiling removes the arithmetic from the discussion, so you can explore an idea properly, say plainly what is not working, and change your mind about a direction you liked last week.

How the batches actually work. We present a set of directions with the thinking behind each one. That is a conversation and not a verdict, and it usually reveals something neither side knew at the brief stage. What we learn shapes the next batch, which is either new territory or a deeper pass on the direction that resonated.

What keeps an uncapped project from drifting. A settled position before design starts, because a project looping on taste is almost always a project with an unresolved strategy underneath it. One person with the authority to decide, since a mark relayed through an intermediary comes back as a list of individually reasonable requests that collectively ruin it. And feedback that names the problem instead of prescribing the fix. Those three do the job a revision limit pretends to do, without making you count.

You will not be shown thirty options, because volume is a symptom of a designer who has not decided anything, and it moves the work of judgment onto you. Behind each batch there are usually sixty or eighty sketches. You see the ones that survived, with an explanation of why.

How the four stages run.

Research. The category, the competitors, and the marks already occupying the space, so we know what to avoid as much as what to aim at.

Explore. Broad sketching, deliberately wide, then narrowed to the directions with genuine potential. This is the stage that decides the outcome and the one clients see least of.

Refine. The chosen direction resolved properly. Optical adjustments, spacing, weight, the lockup system, and testing at every size down to a favicon.

Deliver and protect. Every file, plus the guidelines that keep it consistent once people who were never in the room start using it.

Where a disagreement turns out to be about what the business stands for, that is a positioning conversation. We will pause and have it instead of designing through it, which is frequently the moment a wider consultation earns its place.

You own it outright, transferred on completion, and you receive every format you will ever be asked for. The second part matters more than people expect, because the file you do not have is always the one a supplier needs on a Friday afternoon.

What the delivery actually contains.

Vector master files. The original artwork in an editable vector format, which is the source everything else comes from. Without this you do not really have a logo, you have pictures of one.

Print-ready vectors. EPS and PDF, with type converted to outlines so nothing depends on a font being installed.

Screen files. SVG for the web, plus PNG at multiple sizes with transparent backgrounds.

Every lockup. Horizontal, stacked, symbol on its own, and any secondary arrangement, each as its own file instead of one file you are expected to crop.

Colour variants. Full colour, one colour black, one colour white, and reversed versions for dark backgrounds. Plus values specified in CMYK, RGB, HEX, and Pantone, so what a printer produces matches what the screen shows.

Favicon and app icon files at the required sizes, drawn as their own thing instead of a shrunken logo, since a mark that works at 200 pixels usually needs simplifying at 32.

A single-line or engraving version where the mark will be etched, embossed, embroidered, or cut in vinyl. Almost nobody asks for this until a supplier does, and then it becomes urgent.

On ownership. Rights transfer to you on completion, in writing, including for the final mark and the lockup system. Two exceptions apply, and both are stated in the agreement where you can see them. Unselected concepts remain ours, and any licensed typeface used alongside the mark carries its own terms, which we itemize so you know exactly what you hold and what you are licensing. That distinction matters and it is covered again in the trademark answer.

The files should also arrive organized, named sensibly, and with a short guide to which one to send where. That is part of the guidelines and collateral work, and it is what makes a logo usable by whoever builds your website or your next presentation.

Keep the recognition and fix what is holding it back, which is a different project from replacing it. Most businesses that ask us for a new logo need a careful redesign, and telling them so costs us the larger fee.

When a redesign is the right call. The mark was drawn for print and struggles everywhere the business now appears. It has too much detail to survive a favicon. The files are a mess, or nobody can find the vector. Proportions or spacing were never properly resolved. Or the business has outgrown a literal symbol that describes what it did at the start.

When to leave it alone. People recognize it and associate it with you. It works small. The problem you are actually trying to solve is that your website is dated or your messaging is unclear, both of which the logo will not fix. We audit that honestly, and a fair share of the time the answer is that the mark is fine.

When it needs to go further than the mark. If the position has changed, the audience has changed, or the name no longer fits, a new logo alone will not carry that. That is a rebrand and it should be scoped as one.

How we handle the equity question. First we separate what carries genuine recognition from what is merely familiar. A colour the market associates with you, a shape people can draw, a proportion they would notice changing. Those get protected. Then detail gets simplified, proportions corrected, and the files rebuilt so the mark holds from a favicon to a storefront. Every redesign gets tested beside the old mark at real sizes, on real applications, before anything ships, because a comparison at presentation scale tells you nothing about what a customer will see.

Then the rollout gets sequenced instead of announced. Internal first, then the highest-visibility touchpoints, then the long tail. And if a site rebuild is happening in the same year, the two should be planned together, because launching a new mark on an old site undercuts both.

Nothing stops you, and the reason not to is ownership instead of quality, which is the opposite of what most people assume. The output from a good generator often looks fine. What you may not have is anything you can protect or build on.

On AI generators. Three problems, in order of how much they cost you. First, ownership is genuinely uncertain. Copyright protection generally requires human authorship, and material produced entirely by a generator sits outside that in several jurisdictions, which means you may not hold the rights you assume you hold. Second, these tools produce recombinations of what already exists, so the risk of arriving at something close to an existing mark is real and the tool will not tell you. Third, you get an image and not a system. No lockups, no small-size version, no one-colour variant, no engraving file, and frequently no proper vector master.

On design marketplaces and contests. Cheap, fast, and carrying a specific risk, which is that you are relying on somebody you have never met to have drawn the work themselves. Submissions taken from stock libraries or lifted from other marks do happen, and the liability lands on the business using the logo. Assignment of rights is also frequently informal or absent, which becomes a problem at exactly the wrong moment, meaning a funding round or a sale.

On cheap freelance work generally. Some of it is excellent. What it usually lacks is the category research, the small-size testing, and the file discipline, which is why so many of the marks we are asked to fix came from there and looked fine in the presentation.

Where the cheap route is genuinely reasonable. A side project, a pre-launch placeholder, something testing an idea you may abandon. Spending real money on a mark for a business that might not exist in six months is not prudence, it is procrastination, and we will say so.

Where it stops being reasonable. The moment you intend to register a trademark, raise money, sell through retail, or build recognition you want to keep. At that point ownership and originality stop being abstractions. This is a conversation we have with early-stage founders regularly, and the honest framing is that the cost of redoing it later is several times the cost of doing it once. You can see what the difference looks like in the work.

Usually yes, and the checking should happen before the design is final instead of after, because a mark you cannot register is a mark you cannot fully own. We build screening into the process for that reason.

What actually matters.

Distinctiveness. Descriptive and generic elements are weak. A logo built around a literal picture of what you sell, or a common industry symbol from your category, is harder to protect and easier for a competitor to work around. Distinctive marks are both better design and better assets, which is a rare case of those two things pointing the same way.

The difference between a word mark and a design mark. Registering the name as text protects it in any typeface. Registering the logo protects that specific visual treatment. They are separate registrations covering different things, and most businesses eventually want both. Which one you file first is a real strategic decision.

Classes and jurisdictions. Registration is granted for specific classes of goods and services and in specific countries. Clearance needs to cover the classes you operate in and the markets you sell into, which is wider than the home market on almost every project.

Existing use, and not registrations alone. Unregistered marks in active use can still create a conflict. A search that only checks registers is an incomplete search.

Font licensing, which catches people out. Many commercial typeface licences restrict or prohibit use in a logo or trademark, or require an extended licence for it. This is a genuine and widely ignored problem, and it is one of the reasons custom letterform work is worth it for a wordmark. We convert type to outlines in the final artwork and we itemize any licence that applies, so you know what you own and what you are renting.

Timing. Screening early costs a fraction of discovering a conflict eighteen months in, when changing a mark means new signage, new packaging, new everything, and explaining it to customers.

How we work it. We check exact and close matches, common-law use, and adjacent categories during exploration, so a direction with an obvious problem gets dropped before you see it. We prepare the artwork and documentation your trademark counsel needs, which means their time goes on the legal question instead of on catching up. To be clear about scope, we are not lawyers. Clearance opinions and filing belong with a trademark professional, and we will tell you when it is time to involve one. If you want to talk through where you stand before starting, get in touch.

That is what the refinement stage is for, and it is the part of logo design that separates a mark from a picture. A logo lives in more places than any other piece of design a business owns, and most of those places are hostile.

What gets tested and what each constraint demands.

• Favicon and app icon, at sixteen and thirty-two pixels. The hardest test and the one we apply before showing a concept. Fine detail closes up, thin strokes vanish, and enclosed shapes fill in. Frequently a simplified icon version is drawn deliberately instead of shrinking the main mark.
• Social profile crops. Circular masks and small avatars, which decide how much of a lockup can survive.
• One colour. Fax, stamp, sponsorship sheets, single-colour print, and laser etching. If a mark depends on colour to be legible, it breaks here.
• Reversed on dark. Not simply the same mark inverted. Optical weight changes on a dark background and thin elements need adjusting, which is why a proper reversed version is drawn and not generated.
• Embroidery. Stitching cannot hold fine detail, and there is a minimum size below which a shape becomes a lump of thread. Suppliers will simplify a mark themselves if you do not, and their version will be worse.
• Signage, vinyl, and cut vs printed application. Cut vinyl needs closed shapes with no isolated slivers.
• Vehicle and large format, where proportions and viewing distance change what reads.
• Engraving and single-line application, for anything etched, embossed, or debossed.
• Dark mode on screen, which is now a standard requirement and not an edge case.
• Contrast against the backgrounds it will sit on, which is an accessibility consideration as much as a design one.

The outcome of all that is not one logo. It is a small system, with variants drawn for the conditions they will actually meet, plus rules about which to use where. That is what makes the mark survive contact with the build, your product interface, and whoever your printer is.

A logo on its own drifts, and guidelines are what stop it. Within a year of delivery the mark is being used by printers, partners, sponsors, a new hire, and whoever builds the next presentation, and none of them were in the room when it was designed.

What logo guidelines cover. Clear space and minimum sizes. Which lockup to use in which situation. Colour values for print and screen, including the one-colour and reversed rules. What to do on photography and busy backgrounds. And the misuses that damage a mark fastest, which is stretching it, recolouring it, adding effects, boxing it, or rebuilding the lockup by hand.

The specification we hold ourselves to is short enough to be read and specific enough to settle an argument. A fifty-page document nobody opens protects nothing.

Where a logo genuinely is not enough, and this is the honest part. A mark answers what you look like at the top of a page. It does not answer what your typography is, how photography should feel, what the layout system is, what the voice sounds like, or how any of it behaves across a website, a deck, a sign, and a social post. Businesses that buy a logo and stop usually come back within eighteen months, because the first time three people produce material independently the brand stops looking like one thing.

So the honest sequence looks like this. If you need a mark now and can only fund that, take the logo and the guidelines, and make sure the files and rights are complete so nothing has to be redone. When you are ready, the full identity system extends it into typography, palette, imagery, and layout, and the wider practice settles the position and messaging underneath all of it.

The one thing we will push back on is buying a logo to solve a problem it cannot solve. If the real issue is that nobody understands what you do, a new mark will make that confusion better looking. Say what the actual problem is on the first call and you will get a straight recommendation, including one that is smaller than what you asked for.