Two platforms published what their clients looked for in 2026, and the lists agree more than they should. On Fiverr, searches for "Claude Code" rose 938%, Canva designers 403% and Excel data cleaning 210% between the two halves of the year. On Malt, new tech freelancers leading with JavaScript fell from 40% to under 20% in a year, while briefs for AI agents multiplied by 60. Clients search for tools and outcomes, not competencies. Your profile should name the tool, the result and the decisions you take. Both are the platforms' own numbers, and Fiverr counts searches, not orders.
Surveys tell you which skills people think matter. Search logs tell you what buyers actually typed into a box with a budget behind it. In August 2026 Fiverr published the second kind, and Malt's Tech Trends report from spring covers the same ground for European tech freelancers. Read together, they are the closest thing we have to a shopping list, and they point at a very specific profile rewrite.
What buyers typed into Fiverr
Fiverr's Business Trends Index, June 2026 edition compares global search demand on the platform in two six-month windows: May to October 2025 against November 2025 to April 2026. Two caveats before the numbers. It is Fiverr's own data about Fiverr's own marketplace, published by Fiverr. And it measures searches, not completed orders. A 900% jump in a small category is still a small category.
With that said, the winners cluster.
Named AI tools. "Claude Code" +938%. n8n AI automation +125%. Base44 +95%. Vibe coding +61%. Buyers are not searching for "AI development." They search for the tool they have already read about and want someone to operate.
Named design tools. Canva designers +403%, Shopify graphics and design +348%, Shopify website development +330%, Instagram graphics +149%, logo design +76%. Again, the platform or template is in the query.
Video. Video & Animation overall +278%, AI UGC video ads +265%, YouTube automation for faceless channels +239%.
The unglamorous middle. Excel data cleaning +210%. PDF to Excel +153%. Formatting +71%. Translation +55%. These are exactly the tasks AI was supposed to make disappear, and demand for a human to do them went up. The plausible reading: the tools made the work cheaper and faster, buyers who never bothered before now bother, and they still want a person accountable for the result.
The flat categories. Data overall +3%, data entry typing +10%, generic "Shopify experts" +8%. Notice what these have in common: they are the broad, competence-shaped labels. "Data" is flat. "Excel data cleaning" tripled.
What Malt's tech freelancers stopped writing
Malt's Tech Trends 2026, out in spring 2026 and built on 2025 data from a community of more than 250,000 tech and data freelancers across nine countries, looks at the supply side too. Its most quoted line: about 40% of new tech freelancers highlighted JavaScript when creating their profile in 2024. In 2025 that figure dropped below 20%. Across JavaScript, Python and Java, the share of profiles leading with those languages fell by 20 to 30% in a single year.
That is not JavaScript disappearing from projects. Malt's own demand curve for established languages is described as high but stagnating. It is freelancers deciding that a language is no longer the thing to lead with. What they lead with instead: briefs asking for AI agents multiplied by 60 and searches by 26, n8n project demand grew 14-fold, enterprise demand for Azure OpenAI rose 500%, and the pool of AI-specialised freelancers grew by more than 30%. The fastest-growing job category is not a skill at all. It is a role: interim and fractional tech leadership, up 23%.
Same caveat as above. This is Malt's promotional research about Malt's marketplace, and the report carries no publication date beyond "2026." But the direction matches Fiverr's, from a very different market and a very different price point.
Three patterns, one rewrite
Clients search for tools, then for outcomes, and last for competencies. "Claude Code specialist" outperforms "AI developer." "Canva designer for Instagram" outperforms "graphic designer." The buyer has done the first round of research themselves and arrives with a noun. If that noun is not on your profile, the match never happens.
The AI-execution middle is crowded and getting cheaper. Dorothea Winter, a philosopher at the Humanistische Hochschule Berlin who researches AI and creativity, put the risk in one sentence at Freelance Unlocked:
"My thesis is that AI homogenises the creative middle. If everyone in the middle uses AI, the outputs become more similar, and the thing that defines a creative field, its power to differentiate, gets lost. I consider that a real danger." (Dorothea Winter)
She quoted a study in which AI raised the quality of weaker writers' texts while making them more alike, and left the strongest writers' results unchanged. Translate that to a marketplace: if your offer is "I can produce this with AI," you are competing with everyone who can also produce it with AI, and the search data shows how many of them there are. Her test for your own work habits is worth stealing:
"Where does the AI system enhance me, where is it my sparring partner, and what do I have to do myself, even if it is annoying, slow and less efficient in the moment, so that in the long run I keep my level." (Dorothea Winter)
Victoria Ringleb, managing director of the Allianz deutscher Designer, turned it into a schedule: check two or three times a week whether what you just did was laziness or enhancement, and treat AI as a tool inside your own creative process rather than the thing doing the work.
What grows is orchestration and ownership. Julien Look, founder and CTO of Look Beyond, spent his session on how roles shift inside organisations that use AI seriously. His observation from the teams he works in:
"We are not only seeing developers coding nowadays. That barrier is just not existing anymore." (Julien Look)
Product managers ship features, designers ship clickable UI, QA writes code. And the developers?
"This is exactly where software developers are positioning themselves nowadays. They are basically driving the technical vision." (Julien Look)
Teams shrink as a result. In his words, you no longer need a ten-person team to build a SaaS product; three might be enough. That is the same signal as Malt's fastest-growing category. When execution gets cheap, the person who decides what to build, reviews what came out and owns the result becomes the scarce one. Julien's other point is that the transformation does not end after one project, so he tells consultants pitching AI work to aim for a retention contract measured in years, not a 90-day proof of concept.
"I think there will be only more work created in the future for people in our position." (Julien Look)
What to add to your 9am profile this quarter
Companies on 9am search the way Fiverr's buyers do: by the nouns they already have in their head. Your skills section is where the match is made or missed. Four edits.
- Name the tools, not the category. "Claude Code, n8n, Make" beats "AI automation." "Canva, Figma, Shopify" beats "graphic design." List the ones you actually use for client work this year, not the ones you learned in 2021. The flat Fiverr categories are the competence labels; the tripling ones are the tool names.
- Pair every tool with an outcome. "n8n: replaced a four-person weekly reporting routine with an automated pipeline, 6 hours saved per week." The tool gets you found. The outcome gets you shortlisted.
- Say which decisions you take. If your work includes choosing the architecture, reviewing AI-generated output, or deciding what not to build, write it down. That is the orchestration layer Julien describes and the fractional leadership Malt measures, and almost nobody puts it in a skills list.
- Keep one line that AI cannot homogenise. A sector you know from the inside, a language pair, a regulatory domain, ten years of one kind of client. Dorothea's point in profile form: the differentiator is the part of your work that does not get more similar when everyone has the same tools.
If you lead a tech profile with "JavaScript, React, Node," you are now describing yourself the way 80% of new Malt registrants decided not to. The languages belong on the profile. They no longer belong in the first line. Our skills on the rise 2026 piece has the survey side of the same story, and AI premium or AI discount has the price side: the complex, judgment-heavy AI work is the track that pays more.
What to do on Monday
- Open your 9am skills section and count the nouns. Tools and platforms named: how many? Categories and adjectives: how many? Flip the ratio toward tools.
- Write three tool-plus-outcome lines from projects in the last twelve months. Numbers where you have them.
- Add one decision you own to your headline or first paragraph: what you review, choose or reject on a project.
- Run Dorothea's check on last week's work. Which tasks did AI enhance, and which did you hand over out of convenience? The second list is where your differentiator is eroding.
- Drop one 2021 skill from the first line and move it down. Space at the top is the scarcest thing on the page.
The data says clients arrive with a tool name and leave with the person who owned the result. A free 9am profile with a skills section built that way is how companies across DACH find exactly that person.
Quotes by Dorothea Winter translated from German.
Freelance Unlocked is co-organized by 9am together with Uplink and freelancermap. This article draws on the sessions of Victoria Ringleb and Dorothea Winter, and of Julien Look, at Freelance Unlocked 2026. Watch the full talks above, and join us at the next edition: freelanceunlocked.com.
