Skip to content
Keyword Adv
← All field notes

2025.10.15 · 4 min read

What 800 Upwork jobs teach you about briefs

I have read thousands of project briefs. The good ones are short. The bad ones are long. Here is what to put in a brief that gets work started fast.

I have completed 800 plus jobs on Upwork over 15 years. Top 1 percent expert vetted, which means I see more inbound briefs in a week than most freelancers see in a year. The brief is the first signal of whether a project is going to be smooth or painful. Within the first 100 words I usually know. The good briefs all share the same shape. They are short. They are specific. They tell me what is broken, what they have tried, what they want to fix, and what success looks like. They do not tell me about the company history. They do not list industry awards. They do not include a long preamble about how important this project is. They get to the work. The bad briefs are also recognisable. They are long. They use marketing language to describe technical problems. They explain what they want me to do but not why. They include three different success metrics that contradict each other. They ask for "all the things, properly". They use words like "synergy" or "transformation". They are usually written by someone who is not the person doing the work. If you are writing a brief for a paid media project, here is what to include and what to skip. Include the current monthly ad spend. This is the single most important data point. It tells the agency what tier of operation you are at, what kind of system you need, and whether they can actually help you. If the spend is wrong by 50 percent in either direction, the recommendations will be wrong by the same factor. Include the current performance metric, even approximately. ROAS, CPA, CPL, whatever matters in your business. "Currently doing 3x ROAS, want to scale to 5x" is more useful than "we want to grow". Numbers force specificity. Include what you have tried recently. If you ran a competitor's audit last year and they recommended a Performance Max rebuild, say so. If your last agency exited because of a payment dispute, say that too. Knowing the last 12 months of context saves the new agency from suggesting things that have already failed. Include what success looks like for you. Not what the agency thinks success should look like. If success is "more leads" but you have a 200 lead backlog and no sales team, more leads is not actually what you need. Better quality leads might be. Or fewer leads with higher conversion. The brief should tell me which one. Include the one thing you most need fixed. Not the long list. The one thing. If conversion tracking is broken, that is the thing. If your account structure is a mess, that is the thing. If your landing pages are killing your quality score, that is the thing. The agency can find the rest. They need to know what you have already identified. Skip the company background unless it directly affects the work. The agency is going to look at your website regardless. They do not need three paragraphs of corporate context. Skip the industry awards, the team size, the founding story. None of this changes how I would run your campaigns. Skip the long preamble about how you have evaluated 50 agencies and most are not serious. Just send me the brief. I am also evaluating you. Skip the multiple deliverables stuffed into one project. If you want a tracking audit and a campaign rebuild and a landing page redesign, that is three projects. Trying to make them one project produces a brief that is impossible to scope properly. The other thing the good briefs share is willingness to commit to access. The bad briefs hedge on access. They want me to give recommendations before I have seen the account. This is not how it works. The agency cannot give meaningful recommendations without looking at the actual data. If your brief asks me to estimate without access, the answer is going to be vague, and the project will start with friction. What I do when a brief is bad: I write back asking the missing questions. The response rate is about 50 percent. The 50 percent that respond properly become the projects worth doing. The 50 percent that do not respond were never going to convert anyway. The brief is the cheapest filter in the entire sales process. It tells you who is serious before either side has spent any time. If you are sending briefs to agencies and not getting traction, look at the brief first, not the agencies.

Frank online