Finding the TikTok topics
a garden brand can own —
without making ads.

In 2023 I was on a York University co-op placement at Phoenix Agency. It is worth being precise about that arrangement: a co-op placement is a temporary, university-run work term, so I was a placed student rather than a direct employee of the agency. During that placement, the manager I reported to asked me to come up with content ideas for a TikTok strategy for Environmental Factor — one of the agency's clients, a Canadian biotech brand that makes solutions for protecting a garden or a piece of agricultural land in an environmentally friendly way. The task as handed to me was narrow. I was asked to produce ideas, not content briefs, and certainly not finished posts. What I actually did with it was wider than that, because I thought the only way to produce ideas worth keeping was to build the small strategy that the ideas were supposed to sit inside.

I want to be upfront about something before any of the rest of this. I no longer have the original notes from that co-op assignment, so this is deliberately not a transcript of exactly what I did in 2023. It is a present-day reconstruction — the method as I would approach and solve the same brief today, with the tools and the judgement I would now bring to it. The 2023 placement is the origin of the brief and nothing more; every method step, screenshot and dataset below comes from a fresh re-run I carried out in 2026. I think that is worth stating plainly at the top rather than letting a reader assume otherwise.

The reason I treated a content-ideas task as a strategy task is that TikTok punishes you for skipping the strategy. It is an awareness-heavy platform. People do not show up for over-polished work, and they show up for ad content even less than they do on Instagram. I would say TikTok is really a top-of-funnel platform — organic videos and genuinely helpful content draw people in, and only then do they enter the funnel at all. Content has to feel authentic, and it has to feel like it is starting a real conversation rather than closing a sale.

The argument for top-of-funnel content here is mostly an economic one. If you draw people in with entertainment or education, then once they follow you, you no longer carry the recurring cost of a paid campaign to reach them again — and you have already had a clear opportunity to differentiate the brand before anyone is being sold anything. It also reaches the right people. Awareness content works mostly on the 90%-plus who are still working out which solution they need, or whether they need one at all, and to a lesser extent on people shopping around and comparing. It works much less on people who are already searching to buy, which is the most competitive and most expensive space to fight in.

So before touching a single tool, I had a working hypothesis: the content should start with what customers can do with and around the products, rather than with the products themselves. For a brand that sells garden and soil protection, that pointed fairly clearly at gardening and soil-care content as the place to begin.

The rest of this case study walks the method in order. I have laid it out as a pipeline up front so the whole shape is visible before the detail arrives. Each step below becomes its own chapter, in the same order, and the screenshots from the working log are embedded at the exact point each one informed a decision.

The first concrete thing I did was look at the Environmental Factor website and work out, roughly, what they actually sell. The retail side of the shop is built around three collections — pest control, soil amendments and lawn care — and the commercial side mirrors the same three for professional growers and large-scale operations. So there are three product lines, sold into two sectors. I am not a gardening expert, but the structure of the shop made the brand's own framing clear enough: this is a company that is really about taking care of your soil naturally, by keeping pests out and keeping the lawn and soil healthy.

That told me what they sell. It did not yet tell me what makes them different, and at the awareness and consideration stage, difference is the thing you are actually marketing. You cannot differentiate a brand in content if you do not know what its difference is. So I went to the About Us page, where it becomes clear that the pesticides are biological and free of neonics. Environmental Factor Inc. describes itself as an Ajax, Ontario biotech company developing organic, neonic-free bio-pesticide products.

The word "neonics" was doing a lot of work on that About Us page, and I did not actually know exactly what neonics were. Rather than wave past it, I ran a quick Google search, because if neonic-free is the brand's differentiator then I needed to understand it well enough to build content around it. The AI Overview for neonicotinoids was blunt: they are systemic, water-soluble pesticides that persist in soil and water, killing beneficial insects, pollinators and birds while contaminating ecosystems. A single coated seed can contain enough pesticide to kill over 250,000 bees. They are persistent, mobile, toxic to aquatic invertebrates, and linked to declines in bird and fish populations.

That detour is what gave me the actual USP in a form I could write content around. Stripped of the chemistry, the main selling point of using Environmental Factor products is that you are not killing beneficial insects and destroying the surrounding ecosystem just to produce your garden or protect your soil. That is a sentence a non-expert can understand, and it is a sentence content can be built on.

Environmental Factor sells into both retail and commercial sectors, so B2B content was always going to be possible at some point. But for an initial TikTok content strategy, I considered B2C the better priority, and I decided to focus there first. The reasoning is partly about the platform — TikTok's organic reach skews to consumers browsing for themselves — and partly about being able to test a new strategy cleanly before widening it.

With the sector decided, I sketched the audience pathway I was actually designing content for. A user is browsing TikTok and searches for, or simply engages with, gardening and growing content. Some of that content comes from major brands, and a lot of it comes from dedicated gardening creators — accounts like Creative Explained and Epic Gardening, both of which have built very large followings on exactly this kind of organic, helpful gardening content.

This type of user is almost certainly more likely to respond to, engage with and follow a brand than they are to respond to pure sales advertising — especially if they are not yet sure which product, or even which kind of product, they need. And as they find the content useful, they become more inclined later on to be receptive to a conversion post, or to look the brand up themselves to see whether it has the product they have realised they want. That is the whole pathway: useful content earns the follow, the follow earns the later permission to sell.

This was not a hypothetical for Environmental Factor. Earlier in the same co-op placement at Phoenix Agency I had already produced Instagram posts for the brand, and looking back at them confirmed the pattern. The "4 Benefits Of Having Plants In Your House" post — which I produced myself — is awareness content about the benefits of having plants at all, rather than an attempt to cold-sell a gardening product. Most of the brand's YouTube content followed the same logic: videos about pest, lawn and soil control problems, such as how to get rid of raccoons, rather than sales adverts.

At the awareness and consideration stage you are not really marketing specific products yet — at the awareness stage especially, the job is building reach and getting people to engage with and follow the page. Even so, I took a close look at the product list, because it tells you which topics are even possible to make content around, and which ones will connect back to something the brand can actually sell once a viewer is ready.

The retail pest-control range covers a long list of jobs: grub control; fungus, gnat and thrips control; globe, flea and chinch control; weevil and borer control; slug control; ant control and ant repellent; beneficial nematodes; and pot poppers, which protect against aphids or fungus gnats. On the soil-amendment side there is Turf Maize and its fertilizer for healthier turf, water retention crystals that reduce how often you have to water, Genus MC8 for healthier soil, diatomaceous earth that dehydrates pest insects, and Cropsil for plant nutrients.

The deliberate move here was to record the range by what each product does rather than by what it is. Water retention crystals became "reduce how often you water." Genus MC8 became "get healthier soil." Diatomaceous earth became "control insects naturally instead of spraying chemicals." The one exception I made was pot poppers, which I kept under their product name, because they are a specific, slightly unusual product that people might genuinely be curious about by name.

This is a large list, and the brand could technically make content about every item on it. But especially when you are testing a brand-new strategy, it is far more useful to find out what content people are already searching for — and what is not too competitive — than to assume the whole catalogue deserves equal airtime. That is what the next eight steps are for.

Here is the structural problem with planning TikTok content properly. TikTok does have a trends tool and a Creative Center, but there is no real equivalent of the software you would use in SEO to see specific topic demand — no clean "search volume" figure for a topic the way Google-side tools give you. So a serious content plan for TikTok cannot just read demand off a dashboard. It has to construct the demand signal from several places that were never designed to be combined.

Before doing any of that, I needed somewhere for the signal to land. I settled on three initial content clusters, chosen specifically for non-brand-aware queries — the searches made by people who do not know Environmental Factor exists yet. The three clusters map directly onto what the company sells: pest control, soil amendments and lawn care.

To keep the dataset legible as it grew, I built a single Google Sheet and divided it into those three clusters from the start, so that every term added later — from whatever source — would inherit a known category. I also pre-added the future columns I knew I would need for the per-platform Answerthepublic traffic data, rather than bolting them on later. The sheet began as four rows: a header, then one seed row per cluster, each marked with its source as "cluster itself."

Google Trends is not a TikTok tool, and that is exactly why it goes first. It is free and fast, and before committing time to recursive keyword expansion, it answers one cheap question per cluster: is there meaningful interest behind this topic at all? A topic that is flat at zero on Trends is unlikely to be worth building a TikTok content pillar around, whatever else the other tools say. The pattern per cluster was simple — drop the seed term in, read Interest Over Time, then read the Top queries and Rising queries tables for sub-terms worth promoting into the master sheet.

The first cluster I ran was pest control. Interest in the term itself was reasonable — not sky-high, but with clearly visible, sustained demand. The related-queries tables were busier, but most of the strong terms pointed at pest-control companies — "pest control near me," "best pest control," "pest control services." So I focused only on the terms that could actually work for Environmental Factor at the awareness and consideration stage, and added those relevant ones to the list.

Soil amendments came next, and again showed reasonable interest, which is a good sign for a cluster. Here there were a lot more usable options, and I added them to the list, focusing mostly on terms that could relate either to soil-amendment advice or to products that Environmental Factor actually sells — "organic amendments," "compost," "garden soil amendments," and similar. Lawn care followed the same routine. Many of the lawn-care topics related to contracting a lawn-care company rather than doing it yourself, so I took only the relevant ones, the same way I had with pest control.

At the end of the Trends pass the overall list was not very long yet. That is fine. Trends is the gate, not the verdict. Its job is to confirm each cluster has a pulse and to hand over a small set of credible seed terms — the recursive expansion happens later, with a different tool.

With the Trends pass done, I went to TikTok's own tool — the Creative Center. It is worth being clear about what it is and is not. The Creative Center is built mostly to show data on specific ads and posts, not topic demand the way an SEO tool would. So I did not expect it to replace the demand research. What I wanted from it was narrower: to see whether any of the main hashtags and category terms relevant to the brand were trending, and to pick up format ideas along the way.

I started with the general industry categories. Environmental Factor's products straddle a few of them — Household Products contains a Pest Control sub-category, and Home Improvement contains Interior Design & Decorating. Browsing the keyword popularity inside those categories surfaced a few genuinely interesting terms: "your home," which was both high in volume and rising sharply, alongside "cockroach killer" and "cockroach exterminator." I added "your home," "cockroach killer" and "cockroach exterminator" to the master list, and I also added "home improvement" itself as a new prompt category, since it clearly framed a chunk of relevant demand.

I also looked at TikTok's trending hashtags for more ideas. Inside the Home Improvement category I found a few worth keeping — #garden and #plants, each sitting at around 2,000 posts in the window, and #patio at roughly 850. None of those are surprising for a garden brand, but seeing them register as live, trending hashtags on the platform itself is a useful confirmation that the clusters were pointed in the right direction.

This is the step that builds the demand signal TikTok itself will not give you. The idea is straightforward. A seed term goes into Answerthepublic; every related query that comes back with identifiable traffic becomes the next seed; each of those produces its own children, and so on, until the branches stop returning traffic. The result is a dataset much wider than any single tool would hand you, weighted naturally toward terms that have real demand rather than theoretical demand.

To show the method rather than just describe it, I ran the first term — "pest control" — entirely by hand. Answerthepublic returned a Google search volume of 246K on the seed term, marked High, with a Cost Per Click of $20.64, marked Medium. That figure went straight into the traffic column against what had been an empty seed row minutes earlier, and it was the first real data point in the dataset.

The AI Prompts panel was the first thing I deliberately did not use. It surfaces a long list of natural-language questions people ask AI tools, and the ideas can be good ones — but they are not yet based on actual measured searches, and they carry no traffic data. So I parked them. If that data matures and starts carrying real volume, I would absolutely use it; for now it stays out. The Keywords panel below it did carry volume, so from there I had to weigh which queries genuinely relate to something Environmental Factor sells or provides and still have demand behind them.

The People Also Ask mind map was the most useful single screen in the flow — a visual spider web of question-shaped queries branching off the seed. Here I chose to take all the main terms, because they offered a good mix of awareness and consideration content. "Pest control treatment" is a borderline case: it could refer to a company's service, but it can equally refer to a product, so I judged it worth including. The YouTube and TikTok panels added their own signal — "what is pest control" carried 1.6K of YouTube traffic, and several TikTok-specific terms like "how pest control works" carried real volume too. The longer, informative format that wins on YouTube tends to translate well to TikTok's own longer-video format, so those went into the sheet. The Instagram data I noted as a demand signal even though this is a TikTok strategy, because a topic that draws people on one platform is at least evidence the topic has pull.

Expanding one term manually is fine as a demonstration, but the spider web gets large fast — every term with traffic spawns its own set of children, and doing all of that by hand at scale would have eaten weeks on work that is almost entirely data-shaping rather than judgement. So from this point I set up a Claude project to take the Answerthepublic exports and the manual screenshots of traffic data and write them straight into the spreadsheet. I gave it a fixed set of rules so the automation could not quietly drift.

Once the Claude pattern was working, the dataset grew quickly into a large list of topics — wide enough that, as a searchable database, it could serve many use cases beyond this one project. That breadth is useful, but it also created the next problem. A few hundred candidate topics is too many to commission content against directly. The list needed ranking before it could be acted on.

The plan from here was a manual TikTok audit — watching the ten posts that surface for a query and recording how they perform. But auditing every term in a several-hundred-row dataset is not realistic, so I needed to choose which queries to audit. The audit set I settled on was: all the cluster topics, everything identified through the TikTok Creative Center, everything identified through Google Trends, the top five Answerthepublic TikTok posts, and the top five Answerthepublic Google posts.

To get a ranked shortlist out of the raw traffic numbers, I wrote a small PowerShell script to pull a pure top-ten of the queries with the most identified traffic. The first run came back looking productive and being mostly useless: "garden of the gods" at 368K, "compost bin" at 301K, "patio furniture" and "pest control near me" both at 246K, "plants vs zombies" at 201K, "garden snake," "garden raised bed," "lawn care services," "pest control service." Plants vs Zombies is a video game. Garden of the Gods is a scenic landmark. Most of the rest did not really relate to anything Environmental Factor sells.

So I ran it again, this time excluding the obvious noise — "garden city" came out, and "garden instrument" stayed in its place. Then I changed the script properly, focusing it on the three things the company actually sells: pest control, lawn care and fertilizer. With the ranking scoped to the brand rather than to raw traffic, the top-ten became a list of queries worth the time of a manual audit, and from there I could start the audit itself.

The manual audit produces a lot of rows — every query carries up to ten posts, each with a creator, an engagement figure and notes — and a long spreadsheet is not something a strategy team will actually use. So I turned the audited dataset into a small searchable HTML tool, the TikTok Audit Explorer. It is not meant to be pretty. It is meant to let anyone open it, search a query, sort by engagement, and reach the underlying posts in a couple of clicks — and to keep doing that after I am no longer on the brand.

In its finished version the Explorer covers 129 unique queries and 418 post observations, captured across the 5th and 7th of May 2026, with 314M of total engagement recorded. One column is honestly empty: zero posts carried usable view-count data, because TikTok's search-results cards do not reliably expose views, so the audit recorded engagement and left views blank rather than guessing. Below is a faithful rebuild of the Explorer's top view.

The thing the finished Explorer made obvious is a pattern that runs through the rest of this case study. Most of the demand in the popular queries has to do with topics around what the company's products are used for — composting, soil, gardens, lawns — rather than directly about product features or company information. For a brand trying to sell, that could read as a disappointment. For a brand trying to build awareness and consideration content, it is close to ideal: it is exactly the territory that pulls people into the funnel in an organic way, which is what TikTok rewards.

With the Explorer built, the actual analysis is query by query — watching the top posts that surface for each one and asking what the winners have in common. I worked down the most popular queries in turn. What follows is the read on each, because the patterns are the part that turns a dataset into a content brief.

For the compost query, almost all the content that earned high engagement was do-it-yourself composting tips. That tells you something specific about a gardening brand on TikTok: just featuring the product is not enough. People want to be inspired about how the product can actually be used in real tasks. Composting content, framed that way, could be a strong way to build an organic funnel.

Organic amendments showed the same theme. People like do-it-yourself content that is informational and educational first, and only then shows specific products. The brand could make a video about building a flower bed, and show its product being used to finish the job — which is essentially how grocery stores feature their own products inside the recipes they post online. The product appears because it belongs in the task, not because the video stopped to advertise it.

The garden soil amendments query is where I saw a clear opening to highlight a USP. One of the most popular videos in it was a straightforward "how I changed my garden soil" piece. Environmental Factor could take that exact format and — after a strong hook to engage the viewer — show how its product specifically changes or improves a part of someone's garden soil. The format is proven; the brand just supplies the differentiator.

Organic soil amendments showed a similar opening, with one extra read. People searching this query are likely already asking what the organic alternative is to soil amendments they have used before, which means they arrive half-convinced. But as a criticism of the existing content, almost none of those videos lead with a strong hook. Personally I would open with something like "this changed that about my garden" to draw people in before explaining anything — the demand is there, the hooks are not.

Home pest control was more interesting as a format study than as a topic. Most of the posts were not directly about anything Environmental Factor sells, but the query's popularity flags a real demand for pest-control information, and a recurring format worth stealing: a strong shock hook to pull people in, with some posts literally opening on "don't do this pest control unless you're insane." That shock-hook structure could work well for gardening or, especially, commercial soil pest control.

The #garden query has very high engagement simply because the term is so broad, so my interest there was format rather than topic. The content that wins under #garden is different in kind — short garden skits and timelapse videos largely replace the more informational pieces. Because the niche is so wide, the format I would use is similar skits and timelapses, but ones that ultimately involve the company's products and show the USPs at the end. The related-to-organic-soil-amendments query reinforced the hook lesson again: the posts that performed used a shock hook that names a clear problem and draws people straight into it — the same move you see in effective advertising in fields like healthcare and dieting.

Patio is less relevant to direct gardening or pest control, and the results reflect that. Even so, there is an opportunity to show off beautiful gardens or patios that people made or maintain using the company's products — a kind of product placement, drawing people in with content that has an entertainment angle. The related-to-compost query pointed at a different mechanic: people like longer videos that explain a surprising fact about their do-it-yourself hobby, and the end of a video like that is a natural place for the brand to feature its products or show how to use them for that method.

Termite pest control made the same point again. The winning posts used shock hooks and surprising-fact framing — one of them opening on "you saved the buyer from a future headache" — and the format leaves a clear slot for the brand to feature its products at the end. The tenth query was, again, mostly a format lesson: the company could use it to show how to avoid common gardening problems, opening with a shock hook like "you saved this year's crops by…" and then explaining how the product helps.

The exercise here was never to build a content calendar, let alone produce the content. It was to find the topics and formats worth backing, and to do it with evidence rather than instinct. But the brief did ask for ideas, so the method has to land on some — and based on everything above, these are the ten concrete video ideas I would put forward as actual examples.

There are a few things I want to be straight about. The first I have already said, and it bears repeating: this is a present-day reconstruction, not a record of the original 2023 co-op work. The notes from that placement are gone, so what is documented here is deliberately the method as I would run it today, supported entirely by a fresh 2026 re-run rather than by any 2023 artefacts. The second is scope — this is topic and format discovery, not a content calendar and not an executed campaign, so it tells you what to make and roughly how, but it does not tell you how the resulting content actually performed. That would be the next piece of work, not this one.

The third is that it is deliberately B2C-first. B2B content is possible for Environmental Factor and may well be worth doing, but it was set aside on purpose so the first test could run cleanly on the audience TikTok actually serves. And the audit is a snapshot — TikTok's surface changes fast, so the manual audit should be treated as something to re-run on a regular cadence rather than a permanent answer. None of that undermines the method. The method is the part that survives, because the value here is a process the next person on the brand can re-run and trust, not any single number inside it.

For anyone who wants the raw version rather than the page-by-page excerpts above, the complete 40-page working log is embedded below — every screenshot, every Google Trends read, every spreadsheet state and every audit query in its original order.