Tuesday evening. The kitchen is quiet, and a founder has just sent a 12-slide investor deck that has to be ready by Friday. Most of the work ahead is deciding what belongs on slide three so the numbers land. No tool does that part for you.
Maybe you are reading this with a knot in your stomach. Clients asking whether AI can do the logo cheaper. A studio that has stopped hiring juniors. A day job that pays the bills and still leaves you tense when the car needs repairs.
That fear is fair. It is also why the next year matters so much. Designers who move toward judgment work, decks, systems and production, now are the ones founders call when they need someone who decides. Those who wait keep competing on the tasks that keep getting cheaper.
At the middle of its published range, that one week of evenings on the deck pays something close to a month of median salary. Starting can cost you nothing, or up to $600 if you buy the software and fonts outright, and the first steady $300 a month tends to arrive one to four months in.
Your skills already exist. What changes is what you sell them as, and to whom.
Tonight, open the last deck you built, pick the three slides you would rebuild, and put a fixed price on the rebuild. Write it down where you will see it tomorrow.
You may already feel the squeeze in your own job, and it shows up in the growth rate too. The Bureau of Labor Statistics projects graphic design employment rising from 265,900 jobs in 2024 to 271,500 in 2034. That is 5,700 additional jobs over a decade, a 2.1% increase, against total US employment growth of 3.1% over the same period. Your occupation ranks 457th out of 832 for projected growth. Nothing is collapsing. The field is simply growing more slowly than the economy around it. For you, that means fewer new seats, more people competing for each one, and less pressure on your employer to raise your offer.
BLS says plainly why. Its projections overview lists among the highlights that "the growing adoption of AI technologies, including generative AI tools, and resulting productivity gains are expected to dampen labor demand in a variety of fields, such as sales, design, and administrative support". On design specifically: "Arts, design, media, and communication occupations are expected to be particularly susceptible to productivity effects from generative AI. Graphic designers, for example, can leverage generative AI tools in the initial step of the design process."
Read that last sentence slowly, because your whole strategy sits inside it. The pressure lands on the initial step. So that is where we start.
If you scroll design Twitter or sit in a studio kitchen, you hear one word about AI over and over: "replaced". The projections do not back it up, and seeing why will help you decide what to do next.
Here is the group of occupations BLS singles out as AI-affected, with what it expects to happen to each:
Only three of those shrink. Most grow, slowly. Paralegals are projected to add 600 jobs in a decade across a workforce of 376,000, which is stagnation with a plus sign in front of it.
Stagnation and disappearance are two different problems, and they ask different things of you. If your occupation were disappearing, you would need to retrain. Because it is stagnating, you need to reposition. The work still exists and still pays. What changes is the route to a raise: it stops running through your employer and starts running through what you sell, and to whom.
The table also shows you that "creative" and "safe from AI" have nothing to do with each other. Special effects artists and animators sit at 1.6%, below graphic designers. Meanwhile medical records specialists, a job that sounds like pure clerical automation bait, are projected to grow at more than twice the national rate. What decides your exposure is whether your task can be described in a prompt. How artistic it feels counts for much less.
You are in good company here. Designers have been through this cycle again and again, and the pattern is steady enough to use as a forecast for your own next few years.
Generative AI is the fifth wave, and it behaves like the other four. Each one absorbed a layer of production, pushed the price of that layer toward zero, and left the layers above it intact or worth more. Each one felt like the end of the world to the people whose income sat in the absorbed layer. And each one turned out to be the end of that layer for them, while the occupation carried on.
What feels different this time is the speed and the reach. Earlier waves took a decade to bite and hit one layer each. This one arrived in about two years and touches exploration, production and iteration all at once. The direction is familiar. The squeezed timeline is what makes that 2.1% projection feel uncomfortable when you read it about your own field.
The lesson from the earlier waves is specific, and a little surprising. The designers who did worst competed with the new layer on its own terms: the typesetter who insisted on typesetting, the illustrator who tried to price against stock, the studio that tried to beat a contest on cost. The ones who did best moved up to the decision the new layer could not make. Which of those two people are you being right now, in your own pricing?
BLS says the productivity gain lands on the initial step of the design process. In your day to day, that means exploration and volume production: forty logo directions, thirty social variants, background removal, resizing a campaign across eighteen formats, first-pass layout options, stock imagery, quick mockups.
That work was never the valuable part of what you do. It was the part that was easy to bill by the hour, and those are two different things. Its collapse is why an hourly designer feels the floor moving. If your rate was justified by how long it took you to generate options, generative tools have taken that justification away.
Four things did not get any easier, and they are where your pricing power now lives.
Everything below is built around selling those four things instead of selling your hours.
So now you know which work holds its value and which is being eaten. This next part settles what the work that holds actually sells for.
The gap between what you earn as an employee and what design costs as a purchase is your whole opportunity. Both numbers are public.
At $61,300 a year, your employer is recovering roughly $30 an hour of direct cost for you before overhead. Design agencies publish rates of $100 to $300+ per hour, and freelance rates by experience tier run $15 to $50 on marketplaces, $25 to $50 junior, $50 to $100 mid-level, and $100 to $200+ for senior and specialist work. A mid-level freelance rate is roughly double the hourly value of your salary, and a specialist rate is triple to quadruple it. Which tier would your current portfolio honestly put you in?
Hold onto one more figure, because it describes your actual customer. A typical small business spends $500 to $5,000 on initial design, then $300 to $3,000 per month on ongoing work as it grows. That second number matters most for you, because it recurs, and because it is small enough for an owner to decide without a procurement process.
Start with the smaller end of that range. If one local business keeps you on at $300 a month for social graphics and menu updates, your broadband, your phone and the streaming bills could be covered by an evening or two of design you already know how to do. Offer your last project client a monthly package with a workload that fits that price.
The Highest-Value Hour: Decks
If you want one place to start, make it presentation and pitch deck design. Four reasons stack on top of each other here.
First, the pricing is good and it is published. Entry-level freelancers charge $10 to $35 per slide. Experienced freelancers charge $35 to $150 per slide, or $50 to $150 an hour. Agencies charge $1,500 to $10,000+ per deck. Broken out by job:
| Job | Typical range |
|---|
| 10-slide internal update | $150 - $800 |
| 10-slide sales deck | $1,000 - $3,500 |
| 12-slide investor pitch deck | $1,500 - $7,500+ |
| 20-slide company presentation | $2,000 - $6,000+ |
| Board presentation | $2,000 - $10,000+ |
| Branded presentation template | $1,500 - $8,000 |
| Deck redesign | $500 - $5,000+ |
A 12-slide investor deck at the middle of its range is a week of your evenings for something close to a month of median salary. That sum alone is the reason to look here first.
Second, the stakes are visible, and they justify your price. A pitch deck is attached to a financing round. A board deck is attached to a decision. Nobody argues about a $3,000 design fee on a document meant to raise $2,000,000. Compare that with trying to justify your rate on a social media graphic.
Third, deck work resists automation for a structural reason. A generative tool can lay out a slide. It cannot tell you that slide four is the wrong argument, that the traction chart is technically honest and still reads as a decline, or that the ask should come before the team page for this particular audience. The work is editorial before it is visual, and the editorial part is what your client pays for.
Fourth, the demand is boring and constant. Every company makes decks all the time, and most make them badly. You are looking at a permanent need here, and trends have nothing to do with it.
So stop selling slides and start selling the argument. Ask what decision the deck is meant to produce and who will be in the room. Restructure before you restyle. Then price the deck and never the hour. How fast you work is your own business, and an hourly rate punishes you for being good. Who in your contacts has a raise or a board meeting coming up in the next three months?
The Moat Nobody Wants: Production
Packaging design runs $1,500 to $8,000 freelance and $8,000 to $30,000+ at agencies, and the reason for the premium is unglamorous: packaging can go wrong in ways that cost money.
A dieline has to fold into a real box. A spot colour has to match a reference on a specified substrate. Barcodes have to scan. Regulatory text has to appear at a legible minimum size, and for food and cosmetics the law decides the required content, whatever anyone's taste says. Files go through a printer's preflight and come back with problems. Someone has to sign off knowing that any mistake multiplies by the print run.
Generative tools produce packaging that looks convincing and cannot be built, and that has quietly made production skill worth more to you. You can learn it, too, in a way you can never quite learn pure aesthetics: printers publish their specifications, dielines follow rules, and a few conversations with a production manager will teach you more than a course.
The same logic holds anywhere your output meets a physical or technical constraint: large format and signage, textiles, anything with a bindery step, and on the digital side design systems and component libraries, where "does it actually work at every breakpoint and state" is the same kind of question. Is there a printer near you whose production manager you could buy a coffee this month?
From Hourly to Fixed Scope
The single change that moves your income most is pricing the deliverable instead of the time. Here are three reasons, most important first.
Hourly billing hands every AI productivity gain to your client. Finish a job in three hours that used to take eight and you have just cut your own invoice. With fixed scope, that gain stays with you.
Fixed scope also turns your judgment into an asset. A client buying "a pitch deck, structured and designed, two revision rounds, source files, delivered in ten working days" is buying an outcome and cannot audit your hours. That is exactly as it should be, because your hours were never what they were paying for.
And fixed scope makes you easy for small clients to buy. The $300 to $3,000 monthly ongoing spend of a growing small business gets committed when an owner can say yes to a stated price without approving an open-ended engagement.
A workable productised offer needs a name, one price, an explicit list of what is included, an explicit list of what is not, a fixed number of revision rounds, and a stated turnaround. The exclusions matter more than the inclusions. Scope creep in design arrives as "just one more version", and an unlimited revision count is how your profitable project quietly turns unprofitable.
One caution on positioning. Subscription design services advertise flat monthly rates as low as $399 a month for unlimited requests with 48-hour turnaround, and creative retainers run $2,000 to $6,800+. You cannot beat $399 on volume, and trying to is how a skilled designer like you ends up working like a ticket queue. Those services fight over exactly the commodity production work generative tools are absorbing. Compete on the parts they structurally cannot do: strategic judgment, production accountability, and being a named person who answers for the outcome.
There is a specific opening for you in being the person who directs and checks AI output, while others try to compete with it.
Companies are generating design assets in bulk and running into the second-order problems. Brand consistency drifts across hundreds of generated variants. Generated imagery falls apart against a print spec. Nobody has checked whether the type is legible at the size it will actually appear. Rights and provenance are unclear. There is no system, just a growing pile of files that each look fine on their own.
The services that solve those problems are real, and you can charge for them: building a brand system with rules clear enough that generated output can be checked against it, quality-checking generated assets before they ship, turning AI-generated concepts into production-ready files, and training a marketing team to brief the tools properly.
This positioning has one more advantage. It is honest. If you refuse to touch these tools, you are arguing against a productivity gain your clients can already see, and you will lose that argument. If you use them for the initial step and sell your judgment on top, you are describing exactly the division of labour BLS is projecting. Which of your current clients already has a pile of generated files nobody has checked?
The Rate Ladder
You have picked the work. Now comes the number you say out loud when a client asks, and how you move it up.
Moving from the $25 to $50 band to the $100 to $200+ band is mostly a change in positioning, and much less a change in skill. The steps are fairly consistent.
Pick a deliverable over a discipline. "Graphic designer" puts you up against everyone and every tool. "Investor decks for seed-stage B2B software" puts you up against almost nobody, and that specificity is what lets a stranger decide in ten seconds that you are the right person.
Show your reasoning. A portfolio of attractive artefacts looks the same as a portfolio of generated ones. Three case studies that state the problem, the constraint, the decision and the outcome stand apart, because the reasoning is the thing you are selling.
Charge by project from your very first engagement, even when the first prices feel uncomfortable to say. Moving an existing hourly client over to fixed scope is much harder than starting there.
Raise your prices on new clients only. Existing clients give you stability, and new clients are where you reprice, which spares you the awkward conversation entirely.
Add one neighbouring technical skill. Production and print, or design systems and front-end handoff, or motion. Each one moves you out of the crowd competing on taste and into a much smaller group competing on capability, and taste is the thing that has become cheap. If you had to pick one of those three by Friday, which would it be?
Each rung on that ladder is something concrete at home. The first one covers the phone and the streaming bills. The next pays down the card balance you avoid opening. Higher still, and you can tell a difficult client no without doing sums in your head.
The Retainer, Which Is the Actual Prize
The figure worth building your side income around is the $300 to $3,000 per month a growing small business spends on ongoing design after its initial $500 to $5,000 project. Three or four of those at the middle of the range is a second salary for you, and it is the most stable income open to a part-time designer.
A retainer is also easier to sell than a project, for a reason that has little to do with design. A business owner who has bought design once has learned that every request restarts a hunt: find someone, brief them, agree a price, wait. A retainer removes that friction, and removing the friction is what they are paying for. You are selling availability and continuity as much as craft.
Two structures work at side-income scale. A fixed monthly allowance commits a defined number of hours or requests per month, with a stated turnaround and unused capacity that does not roll over. It is simple to sell. The risk is that a slow month feels expensive to your client and a heavy month feels underpaid to you. A defined deliverable set commits you to specific outputs each month, such as a set number of social assets plus one larger piece. That prices more predictably for both of you and is easier to defend when the client's needs spike.
Three conditions keep a retainer profitable and stop it becoming a trap. Cap the scope in writing, because "ongoing design support" with no boundary turns into an unpaid employment relationship. Set your turnaround honestly around the fact that you have a day job: a promised 24-hour response you cannot keep damages the relationship faster than a stated three-day one ever would. And review the price every six months, because retainers drift. Your client's needs grow, your rate stays put, and after two years you are working a second job at your first-year rate. Could you hold a three-day turnaround on a week when your day job is on fire?
The best retainer clients share a profile. They are past the founding stage and producing regularly, small enough that an in-house designer makes no sense, and already spending on design in a scattered way. An in-house designer costs $4,200 to $9,500 a month loaded. That is the ceiling your retainer sits under, and knowing it gives you a calm answer in any pricing conversation.
Three or four retainers at the middle of that range is the second salary this section is about. Once your own costs are paid, that money could let you take over your parents' winter utility bills before they ever have to ask, paid for by brand work for a café and a local gym. Start with one. Ask the client you already design for whether they would like ongoing support.
Where the First Three Clients Come From
Your pricing is settled. None of it counts until someone on the other end says yes.
The honest answer is that your first clients will rarely come from a marketplace, and I want to be blunt about that, because a lot of evenings get lost there.
People who already know your work. Former colleagues, people who left your employer for other companies, the marketing manager you used to hand files to. These are the warmest leads you have, because the trust question is already settled, and trust is most of what they are buying. One message saying what you now do and who it is for, sent to twenty of them, will beat months of polishing a profile. Who are the first five names on that list for you?
Agencies and studios that need overflow. Studios turn work down or hit capacity all the time, and a reliable specialist who can take a deck or a packaging job without supervision is genuinely useful to them. You will earn less than with direct clients because they hold the relationship, and in return they do the selling. If you have a day job and no time to chase leads, this is often your best first channel.
Neighbouring professionals with the same client. Copywriters, developers, brand strategists, fractional marketers and accountants all sit next to your buyer and get asked for design recommendations. Two or three two-way referral relationships here bring steadier work than any amount of advertising, and they cost you nothing to build.
Narrow, visible specificity. A single page that says exactly what you do for exactly whom, with three case studies showing your reasoning, turns strangers into clients far more often than a general portfolio. Specificity lets someone decide in ten seconds that you are the right person, and generalists never get those ten seconds.
What reliably fails at this stage is competing on marketplaces where buyers sort by price, or posting work into a feed and waiting. Both feel productive. Neither reaches the buyer who has a $3,000 problem.
Every week you wait is another week of portfolio pieces nobody sees and clients picking someone braver. You already have the work. Write the single page that says exactly what you do for whom tonight, and send it to one founder you know before the weekend.
Rookie Mistakes
Competing with the tools on volume. Offering more concepts, faster, cheaper. You are racing something with no marginal cost, and every hour you spend in that race teaches your client to value throughput.
Pricing from your salary. If you earn $61,300 and work out an hourly rate from it, you land around $30 and anchor yourself low for good. Freelance rates carry no employer benefits, no paid holiday, unbillable admin and sales time, and self-employment tax. The mid-level $50 to $100 band exists partly to cover those costs, so it is barely a pay rise over $30 at all. What would your hourly number be once you add those costs back in?
Unlimited revisions. The most common way a project makes money on paper and loses it in practice. Offer two rounds, clearly defined, with a stated price for any more.
Selling logos. A logo at $300 to $2,500 attracts the most opinions per dollar of anything on the price list, because everyone in your client's company has a view and none of those views are held back by evidence. Brand systems, decks and packaging have narrower approval loops and better economics for you.
No written rights terms. Design changes hands without anyone talking about ownership until it suddenly matters. Write down what you transfer, when it transfers, whether it depends on final payment, and what is licensed as opposed to assigned.
Treating the day job as the enemy. As a salaried designer you have something a full-time freelancer lacks: the freedom to decline bad work. That is your single biggest advantage when raising rates, and it deserves to be used on purpose instead of rushed past.
Gotchas Worth Knowing
Your employment contract may own your evenings. Many design roles include intellectual property assignment clauses broad enough to cover work you make outside working hours, and some include non-compete or non-solicitation terms. Read your contract before you take a client, especially one in your employer's sector. When did you last actually read yours?
Generated assets carry unclear provenance. If you use generative tools in client work, know what your tool's terms say about commercial use and indemnity, and say in your own contract what you used. A client who finds out later, in a dispute, is a far worse conversation than a client you told upfront.
Client-supplied assets are a liability. Fonts without licences, stock images without rights, a competitor's photograph pulled from a website. You are the professional in the room, and a licensing problem you shipped becomes partly yours. Ask for provenance in writing.
Marketplace pricing is a trap with a floor of $5. Contest platforms and marketplace tiers at $15 to $50 an hour rarely lead you to $150 an hour, because the buyers there choose on price and will not follow you up. Referrals and a specific niche get you there; marketplaces mostly get you volume at the rate you started at.
Print mistakes have no undo. The first time a job goes to press wrong, you will remember it for years. Order a hard proof, get your client to sign off on the proof itself instead of a screen, and write that step into your process.
Deck work is deadline-shaped. Pitch decks arrive with a board meeting or a fundraise attached, so the timeline cannot move and the content always lands late. Price rush turnaround explicitly and hold the line, or this category will eat your weekends indefinitely.
Behind the Scenes: A Deck Engagement
Here is how one of these usually goes, and it involves less design than you might expect.
A founder sends you twenty-two slides in a shared deck, mostly text, with a note saying they pitch on Thursday week. You read it through and still cannot tell what the company does.
The first call is an hour, and none of it is about visuals. What is the ask. Who is in the room. What do they already believe. What is the one thing they must remember. The founder answers that last question three different ways. That is your real finding, and the deck cannot be fixed until it is resolved.
Then you restructure, in outline, with no design at all. Twenty-two slides become fourteen. Two slides get merged, four get cut, and one gets added because the story has a hole where the business model should be. You send this as a plain list, and it is the most valuable thing you will deliver in the whole engagement.
The founder pushes back on the cuts, which is normal. You restore one, because they were right and it mattered to their audience. Two stay cut.
Then comes design, which is the fast part once the argument is settled, and where generative tools genuinely help you with imagery and first layouts.
Then the chart problem, which is always there. The traction number is real, and the default chart makes it look like a plateau. Fixing that honestly, by changing the axis or the timeframe or the comparison without misrepresenting anything, takes longer than five slides of layout. It is the most consequential decision in the file.
Then revisions, a rehearsal where the founder discovers slide nine falls flat when spoken aloud, and a final file plus an editable version.
Your time is not spread evenly. Roughly a third goes on structure, a third on design, and a third on the two or three decisions that decide whether the document does its job. Billed by the hour, that engagement is unremarkable. Billed at $1,500 to $7,500 for an investor deck, it is the best-paid work most designers can reach. How would you price this exact job if it landed in your inbox tomorrow?
When to Stop Doing Both
By now your side income is real and your job is still there. This is where you work out how long holding both is worth it.
Most guides push you toward quitting. The projections argue for more patience.
A salaried job in a stagnating occupation is still in a stagnating occupation, so treating it as safe forever would be a mistake. But your side income has to clear a higher bar than matching your salary, because a $61,300 job is worth more than $61,300 to you. It includes your employer's share of payroll taxes, any health coverage, paid leave, equipment, and the comfort that a bad month costs you nothing. Replacing it means covering all of that plus self-employment tax plus the unbillable hours you spend selling. In practice, your gross self-employed revenue needs to sit well above your salary, and merely equal is too thin.
Three signals tell you the switch is real and more than hopeful. You are turning work down for lack of hours, again and again, over several months and beyond one good quarter. Your income comes from more than one client, so losing a single relationship cannot reset you to zero. And you have a pipeline you can describe: you know where the next client comes from, instead of hoping the last one refers you.
Two signals tell you to wait. Income concentrated in one client is a job with worse terms and no notice period. And a run of good months driven by one unusual project is a fluke, however much it feels like a trend. If your biggest client left tomorrow, how many months could you cover?
The strongest reason to stay employed longer than feels necessary is the one from earlier in this guide: with a salary behind you, you can decline underpriced work. That one ability is what moves you up the rate ladder, and it vanishes the moment your rent depends on the next invoice. Plenty of designers leave the month their freelance income first matches their salary, then spend two years taking work at rates they had already outgrown, because they can no longer afford to say no.
Used on purpose, your day job funds your repositioning. That is a better use of it than treating it as something to escape.
Hold the top of this guide's income range, near 7,500 a month, long enough to cover the payroll tax and benefits a $61,300 salary quietly includes, and leaving becomes simple arithmetic. You hand in your notice with months of runway saved and a roster of retainers that already pay. Until the numbers say so, your day job keeps buying you the patience to turn work down.
Where This Goes Next
Here are four judgements about where this is heading. They are reasoning from what you can already see, and they are offered as reasoning, so treat them with the care you would give any forecast. Each one changes what is worth learning next.
The middle of the market thins first. Commodity production work is going to generated tools and $399 subscriptions. Strategic and accountable work holds its price. The uncomfortable spot is the competent generalist charging $50 an hour for execution, and that is where most employed designers sit today, which is what the 2.1% projection describes. Is that where you are sitting?
Employment stagnates while engagement holds. BLS expects 5,700 net new design jobs over a decade. Freelance and project demand is a separate market, and small business ongoing spend of $300 to $3,000 a month does not appear in an employment projection at all. Expect the gap between salaried and self-employed design income to widen for the people who make the switch on purpose.
Production and systems knowledge gains value. As generation gets cheaper, the constraint moves to whether output actually works: on press, at every breakpoint, within a brand system, under regulation. Designers who can answer that will be scarcer relative to demand than designers with good taste.
Brand systems get bought earlier. A company generating hundreds of assets needs clear rules sooner than a company producing twelve assets a year. Systems work that used to be a late-stage purchase becomes an early necessity, and that is a genuinely new market you can step into.
Titles stop describing the work. The useful description shifts from a discipline to an outcome, and you can already see it in how the best-paid work is sold. "Graphic designer" describes what you can make. What you get paid for is what you can be trusted to decide.
Companies are choosing their design partners for the AI era right now. Whoever sets up their brand system or rebuilds their deck becomes the person they call next time. Wait a year and those relationships belong to designers who moved first.