How to research a client's niche fast when you have never worked in it
You signed a client in a field you know nothing about, and the first content plan is due this week. You do not need to become an expert. You need to find out what this audience has already rewarded.
Short answer. To research a new client niche quickly, start from evidence and not opinions. List 10 to 15 accounts around the client's size, find the posts that beat each account's median, and sort them by topic and format. Add the exact language buyers use from comments, reviews and forums. Then interview the client for one hour about objections and stories. In about a day you can turn that into a one-page brief and a first batch of content ideas.
As one marketer asked in a thread, "How do this when your client or boss don't give you hours to research the topic in your scope?" The answer is a fixed process with a time limit. The steps below assume about one working day. Each has a time box, and it is fine to stop when the box runs out.
Hour 1: Map the accounts
Do not start with the biggest names. Build a list with three kinds of account.
- Direct competitors. Businesses that sell what your client sells, ideally in the same market. Ask the client who they lose deals to.
- Similar-sized creators. Individuals in the field with an audience near your client's size. These are your best evidence.
- Adjacent accounts. Those who speak to the same buyer about something else. A wedding photographer can learn from wedding planners.
Aim for 10 to 15 accounts in a shared sheet, with the platform and rough follower count. Remove any account that clearly runs on paid reach or a celebrity founder, since you cannot reproduce that.
Hours 2 and 3: Find what beat its own baseline
For each account, collect recent posts and score them against that account's median, as in how to find outlier posts. Note anything at twice the median or more. Copy these into a sheet with five columns: link, score, topic, format and opening line.
This is the main saving from using a tool. By hand, 15 accounts takes hours. XCut AI does the scoring for you across Instagram, TikTok, YouTube and Facebook, and you can pull in several accounts at once, but the logic is the same if you are working by hand.
When you finish, sort by topic. You are looking for topics that appear across several accounts. One account's outlier might be luck. Five accounts with outliers on the same topic is close to a fact about this audience.
Hour 4: Collect the audience's words
Now find out how the buyers actually talk, since you are not one of them. Spend an hour reading and copying exact phrases.
- Comments on the outliers. What people say, ask and argue about.
- Reviews. Reviews of the client and its competitors, especially the three-star ones. They tend to be the most specific.
- Forums and Reddit. Search for the problem and not the product.
- The client's inbox. Support tickets, sales emails and call notes, with personal details removed.
Sort what you find into three lists: questions people ask, results they want, and doubts they have. Keep their exact words. These lists become your hooks.
Hour 5: Interview the client
Go in with your evidence, which makes this a much better conversation. Asking "what should we post?" usually gets you a list of the client's favourite topics. Ask these questions instead.
- What do customers misunderstand before they buy?
- What do you find yourself explaining on every sales call?
- Why do people say no?
- What do you believe that others in your industry would argue with?
- Tell me about your best customer outcome. What exactly happened?
- What can we not say? Ask about compliance, claims and past problems.
- Which competitors do you admire, and which annoy you?
Record the call and transcribe it. The stories and opinions in that transcript are the raw material no competitor has, and they are also your voice reference for the brand.
Hour 6: Write the one-page brief
| Section | What goes in it |
|---|---|
| Audience | One real buyer, in their own words |
| Three topics that win | From your outliers, with links |
| Three formats that win | From your outliers, with links |
| Hook language | Exact phrases from comments and reviews |
| Client's edge | The opinions and stories only they have |
| Off limits | Claims and topics to avoid |
| First tests | Six content ideas, each tied to evidence |
Each claim in the brief should link to its evidence. It changes the client conversation. You are no longer saying "we think you should do this." You are saying "here are six posts at your size that beat their own average, and here is why ours should follow them."
Hours 7 and 8: Draft the first batch
Pick the two strongest topic and format combinations and draft about six scripts. Use outlier structures with the client's stories inside, as in turning a viral reel into your own script. Mark each one with the evidence behind it, and have the client check accuracy. In a technical or regulated field, this review is what protects you from your own lack of knowledge.
Regulated and specialist niches
Health, finance, law and similar fields need more care. Do not treat a competitor's claim as proof that the claim is allowed, because they may simply not have been caught. Ask the client who signs off on content, and agree on this before you draft anything. Stick to explaining and away from promising results. A good rule is that the client approves any factual claim you could not defend yourself.
Running this for many clients
If you make 30 or more short videos a month across clients, the process has to become a system. Three things tend to break first.
- Voices blur together. Keep a separate voice file and brief per client, and never reuse them across accounts.
- Research goes stale. Repeat the outlier step monthly. It takes much less time once the account list exists.
- Writers wait on research. Keep an idea bank per client, each idea linked to its evidence, so writing never starts from a blank page.
A workable monthly rhythm for 30 videos: one research session per client, one batch writing session, one client review, then record in batches. This is how a small team stays ahead without hiring a dedicated writer for each account, though someone still needs to edit for accuracy and voice.
XCut gives each client a separate canvas and memory, so brand voice and research stay apart. See XCut for agencies and the best tools for small agencies. The free niche validator and competitor gap finder are handy for a first look before a pitch.
A one-page brief you can copy
Clients rarely read long research decks. Use this template, keep it to one page, and put the evidence in an appendix for anyone who wants it.
| Section | What goes in it | Length |
|---|---|---|
| Who we are talking to | One specific person, in plain words | 2 lines |
| What they want | Their words, quoted from comments or reviews | 3 quotes |
| What stops them | Their doubts, also quoted | 3 quotes |
| What works in this niche | Outlier patterns, each with a link | 3 patterns |
| What is overdone | Angles every account already uses | 2 lines |
| Our gap | What the client can say that others cannot | 2 lines |
| First tests | Five post ideas with the pattern each one tests | 5 lines |
| Rules | Claims to avoid and words the client will not use | 3 lines |
The "Our gap" row is the one clients care about. Anyone can list what competitors do. Telling a client what only they can credibly say is the reason they hired you.
Questions to ask the client
The one-hour interview works best when you get stories, not opinions. Clients tend to describe their business the way their website does. Use questions like these to get past that.
- What do customers say in the first two minutes of a sales call?
- What is the most common thing people get wrong before they come to you?
- Who is a bad fit, and how can you tell?
- What do you believe about this industry that competitors would disagree with?
- What do you find yourself explaining over and over?
- Tell me about a customer who nearly did not buy. What changed their mind?
- What are you not allowed to say, legally or by your own rules?
- Which competitor do you respect, and why?
Record the call if the client agrees. The answer to "what do you find yourself explaining over and over" is often a month of content, already in the client's own voice.
When you do not get a full day
The eight-hour plan is the ideal. Real work is often tighter. One person asked: "How do this when your client or boss don't give you hours to research the topic in your scope?" Another wrote: "Diving into a new niche and researching about it is soo time-consuming but if you can earn from it eventually, then it's worth it!" If you only get a short window, keep the steps that produce evidence and cut the rest.
| Time you have | Keep | Cut |
|---|---|---|
| Two hours | Three accounts, their outliers, 20 audience quotes | The interview and drafting |
| Half a day | Five accounts, quotes, a 30-minute client call, a short brief | The first batch of drafts |
| A full day | The whole process in this guide | Nothing |
Tell the client what the shorter version leaves out. A two-hour read gives you a starting point, and it should be labelled as one.
Checking your work before it reaches the client
It is easy to sound sure in a niche you learned this morning, which is the danger. Run these checks before you send anything.
- Every claim about what works has a link to a real post.
- Every quote is from a real person and is word for word.
- Nothing in your drafts makes a health, money or legal promise the client cannot support.
- You have asked the client or an expert to read for mistakes in terminology.
- You have said which of your conclusions are guesses.
That last point earns trust. A note saying "this is based on five accounts over three months, so we will test it" reads as more professional than a confident claim that turns out to be wrong in week two.
What not to do
- Do not rely on general AI knowledge of the niche. It gives you plausible content that every other agency also has. Evidence from this month beats a summary of the past.
- Do not copy the market leader. Their reach comes from things your client does not have.
- Do not skip the client interview. Their real stories are what make the content impossible to copy.
- Do not over-research. One day is enough for a first batch. Real results will teach you more than a second week of reading.
Frequently asked questions
How long should niche research take?
About one working day for the first pass on a single client, then a shorter monthly refresh. Real post results will tell you more than extended research.
How many competitors should I look at?
Ten to fifteen accounts is enough to see what repeats. Mix direct competitors, similar-sized creators and adjacent accounts.
How do I sound credible in a niche I do not know?
Use the audience's own words from comments and reviews, and the client's real stories. Have the client check every factual claim.
What should I show a client in the first week?
A one-page brief with linked evidence and six ideas. It shows you did the work and gives the client something specific to react to.
Can AI do this research for me?
AI can sort and summarise what you collect. Without current post data it cannot tell you what is working now, so it needs real evidence to work from.
What if the client has no social history at all?
Rely on competitors and audience language, and treat your first month as testing. You will have no baseline, so build one from the first 10 to 20 posts before you judge what is working.
Should I share the research with the client or keep it internal?
Share the one-page brief. It shows why you are recommending each post, which makes approvals faster and makes your work harder to replace.
References
See which of your posts beat your average
Paste your handle at xcut.ai/instagram for a free scan with no account, or start a trial to research any niche.
Start your 7-day trial$5 for 7 days · 1,000 credits · cancel anytime