Week 1
Define the territory
What you sell, what you can credibly speak about, and therefore what is worth watching. Without this, everything looks interesting.
AI-assisted workflows · Trend spotting · Sydney
Trend spotting
Every topic has a window where demand exists and nobody good has written about it yet. The window is usually a few weeks, and almost nobody is watching for it because watching is dull and continuous.
Grab the snowball. Give it a spin, and watch it grow as you scroll. Just like the signal.
6sources watched continuously
1short window per topic
Machine watchesa person decides what matters
Weeklybecause monthly is too late
The short answer
What is trend spotting, and why does timing decide it?
Trend spotting is watching for questions and terms that are rising before they are competitive, so you can publish while the topic is still winnable. It is a monitoring discipline rather than a creative one, and it fails from inattention rather than from lack of skill.
Timing is almost the whole game. A topic with rising demand and no good answer is cheap to win and stays won. The same topic six months later has four thorough articles from sites with more standing than yours, and the cost of competing has multiplied.
We use tools to watch continuously and a person to decide what is worth acting on, which is the division of labour that actually works: machines are good at noticing and bad at judging. Because our Snowball SEO platform automates the search heavy lifting other agencies bill by the hour, the watching costs you very little. It sits inside our wider AI-assisted workflows service.
The difference
How long is the window actually open?
Weeks, usually. Long enough to act if somebody is watching and far too short to catch in a quarterly planning cycle, which is how most content is planned.
What is worth watching
- Question volumes rising against a flat baseline, which is the clearest signal
- New terminology entering an industry, before anyone has defined it well
- Regulatory and legislative changes with a known commencement date
- Questions appearing in your own sales calls before they appear in search
What is noise
- Anything already in the trade press, which means the window is closing
- Seasonal spikes mistaken for growth
- Terms with volume and no commercial relevance to you
- Whatever a competitor published last week, which is a lagging indicator
Acting on what is found is our article writing work; this page is the watching that decides what to write.
The cost
What drives the price of trend watching?
Breadth and frequency. Watching six sources weekly is cheap and continuous; watching everything daily is neither, and rarely produces better decisions.
A watch brief
Monthly
A monitored set of topics and terms with a monthly read on what is moving and what is worth acting on.
Weekly watch
Weekly
The same, weekly, which is the cadence that actually catches windows rather than reporting on ones that have closed.
Watch and act
With production
The watching joined to the writing, so a window found on Tuesday is published against rather than added to a backlog.
We quote by breadth and cadence after a free consult. For the argument about using tools without losing voice, see our workflow guide.
What moves the number, in order
- How many topics and terms are watched
- Cadence, since weekly catches windows and monthly reports on them
- How many sources, including ones with no API that need a person
- Whether your own sales and support data is included, which is the best source
- Whether findings come with a recommendation or just a list
- Whether production is joined to the watching
The best source is your own inbox
Every business already receives early signal and discards it. The questions arriving in sales calls and support tickets this month are the searches of six months' time, and nobody is aggregating them.
It costs nothing to collect, it is specific to your market rather than generic, and no competitor has access to it. It is the single most underused input in content planning.
So we start there before any external tool, and where a client cannot easily give us that data, fixing the collection is usually worth more than the watching.
The scope
What you receive
A short read you can act on, not a dashboard nobody opens.
The watch list
The topics and terms under observation, agreed with you, and revised as things move rather than set once.
A weekly read
What moved, what it probably means, and whether it is worth acting on. Short enough to be read, which most monitoring output is not.
Recommendations
Not just what is rising, but whether you specifically should act, given what you sell and what you can credibly say.
Question capture
A simple way to collect what your sales and support teams are being asked, since that is the earliest signal available.
What we ignored
The things that moved and were not worth acting on, with reasons, because a watch list that only reports positives is not a judgement.
A handover note
Enough for your team to act without us, because a monitoring service that creates dependency is not doing you a favour.
The return
Where does a topic sit when you find it?
Somewhere on a line between nobody has asked this yet and everybody has answered it. Only a narrow band is worth acting on, and it is not the middle.
Most findings should be discarded
A trend watch that recommends acting on everything it finds is not watching, it is generating work. The value is in the discarding, and a good weekly read has more rejections than recommendations.
Discarding well means knowing what you can credibly say. A rising topic you have no genuine expertise in is not an opportunity, it is a way to publish something shallow with your name on it.
So each finding is judged against what you actually know and sell, which is why the recommendation matters more than the detection.
The process
How trend watching runs
A fortnight to set up, then a weekly rhythm that takes very little of your time.
Week 1
Wire the sources
Search and question data, industry sources, and a way to capture what your own sales and support teams are being asked.
Week 2
First read
A baseline of what is currently moving, which is also a useful audit of what you have missed in the last year.
Weekly
The read
What moved, what it means, what to act on and what to ignore, with reasons on both sides.
Monthly
Revise the list
Topics that have saturated come off, new territory goes on. A static watch list stops working within a quarter.
Quarterly
Check the calls
Which recommendations were acted on, and what happened, because a watch service that is never scored drifts into noise.
The brief
What to ask before you buy trend watching
Monitoring is easy to sell and hard to judge, because the output looks the same whether or not it is any good.
How often will we hear from you?
Weekly at least. Monthly reporting describes windows that have already closed, which is history rather than intelligence.
Will you tell us what to ignore?
A good read discards more than it recommends. One that finds an opportunity every week is generating work rather than exercising judgement.
Are you using our own data?
Sales and support questions are the earliest and most specific signal available, and they cost nothing. A service that only uses external tools is leaving the best source unopened.
Can we act on this without you?
The read should be usable by your own team. Monitoring that only creates work for the supplier who found it is a conflict of interest worth naming.
We are happy to answer all four before you commit. On using tools without losing voice, see our guide to AI-assisted workflows.
Start here
Ready to see it early?
Tell us what you sell and what you can credibly speak about. The free consult is a working session, not a sales call, and you leave with three things whether or not you book us.
- A first read on what is currently moving in your territory
- The two topics we think are in the window right now, with reasons
- A fixed price by breadth and cadence, with your own data sources included
Good questions
Trend spotting FAQs
What is trend spotting for content?
It is watching for questions and terms that are rising before they become competitive, so you can publish while a topic is still winnable. It is a monitoring discipline rather than a creative one, and it fails from inattention rather than from lack of skill.
How long is the window on a topic?
Usually weeks. Long enough to act on if somebody is watching continuously, and far too short to catch in a quarterly planning cycle, which is how most content calendars are actually built.
How much does trend watching cost?
We quote by breadth and cadence. What moves it is how many topics and terms are watched, how often, how many sources need a person rather than an API, and whether findings come with a recommendation or as a list.
Do you use AI for this?
For the watching, yes: noticing changes across many sources continuously is exactly what machines are good at. The judgement about whether something is worth acting on is made by a person, because that requires knowing what you can credibly say.
What is the best source of early signal?
Your own sales calls and support tickets. The questions arriving there this month are the searches of six months' time, they are specific to your market, and no competitor has access to them. Almost nobody aggregates them.
Should we act on everything you find?
No, and a read that recommends acting on everything is generating work rather than exercising judgement. Most findings should be discarded, and the reasons for discarding are as useful as the recommendations.
What if the topic is outside our expertise?
Then it is not an opportunity. A rising topic you cannot speak about credibly is a way to publish something shallow under your own name, which costs more trust than the traffic is worth.
How do we know if it is working?
Score the calls. Which recommendations were acted on, what happened, and which discarded topics turned out to matter. A monitoring service that is never scored drifts into noise within about two quarters.