Skip to main content

How to find trending products before they peak

By

Trend discovery tools are lagging indicators that sell you the top of the market. By the time a product charts on a trend aggregator, the early adopters have already built their distribution channels and the cost of acquiring a customer has doubled. You cannot build a sustainable business by chasing a spike. You build it by separating temporary viral noise from compounding, multi-year demand.

The second year reveals the truth

A viral spike in month one tells you nothing about market viability. The second year of a trend matters entirely more than the first. When a product launches and goes viral on social video platforms, sales explode. This is the spike. Then the algorithm moves on, the influencers stop posting, and the initial wave of buyers realizes they do not actually need the item. The spike collapses.

What happens in month thirteen is the actual demand signal. If the baseline sales level in year two is higher than the baseline in year one, you have a real trend. If it returns to zero, you have a fad. You must ignore the peak and focus exclusively on the floor.

Doing the math on search history

To find that year-two baseline, you need historical search volume data. Look at a three-year chart for your target product category. Do not look at the peaks. Draw a horizontal line through the lowest search volume month of year one. Draw another horizontal line through the lowest month of year two. Calculate the percentage difference between those two floors.

Consider a specific example. If a search term hits 100,000 searches in November but drops to 10,000 in March, the floor is 10,000. If the next year hits 150,000 in November but the March floor drops to 8,000, the trend is dying. The peak grew, but the underlying baseline shrank. The product is becoming more seasonal and less essential.

Conversely, if the peak stays flat at 100,000 but the March floor rises from 10,000 to 18,000, you have an 80 percent increase in baseline demand. That is a product worth building. The people searching during the off-season are your core market. They are buying the product because they need it, not because it is trending.

Calculating the decay rate of a spike

When a product experiences a viral event, the search volume and sales will drop eventually. The speed of that drop is your strongest indicator of long-term viability. You need to calculate the decay rate. Look at the peak month of search volume, then look at the volume three months later.

If a product hits 500,000 searches in July and drops to 50,000 in October, it has a 90 percent decay rate over three months. This is a severe drop, indicating the product was likely a novelty item or heavily dependent on a specific social media trend.

If another product hits 500,000 searches in July but only drops to 300,000 in October, the decay rate is 40 percent. The market is retaining a massive portion of the new audience. A slow decay rate means the product is crossing the gap from early adopters to the mainstream market. People are seeing the product, buying it, and then recommending it to their friends offline. Word of mouth creates a much slower decay curve than algorithmic social video feeds. You want to build products in categories with a 90-day decay rate of less than 50 percent.

Decoding marketplace sales ranks

Major ecommerce marketplaces assign a sales rank to every product. A rank of 1 means it is the highest-selling product in that category. Most sellers look at the current rank, see a low number, and decide to source the product. This is a mistake. A product can hit rank 50 because a large influencer mentioned it yesterday. Tomorrow, it will be rank 5,000.

You must track the average sales rank over a 90-day period and compare it to the previous 90-day period. You want a product where the average rank is slowly improving, not wildly fluctuating. If a product averages a rank of 15,000 in Q1 and 12,000 in Q2, the sales velocity is increasing steadily.

To estimate the actual sales volume, you can use category-specific multipliers. For example, in a major home goods category, a rank of 10,000 might equal 150 sales per month, while a rank of 5,000 equals 400 sales. If you track a competitor moving from 10,000 to 5,000 over six months without running massive discounts, you have found a rising trend before the market saturates.

Spotting inventory stockouts

When you are tracking average sales ranks, you will occasionally see a rank drop from 5,000 to 150,000 overnight. Inexperienced sellers assume the product died. Experienced sellers know this is a stockout. The competitor sold out of their inventory, and the marketplace algorithm suppressed their listing.

A stockout is a massive demand signal. It means the competitor underestimated the market demand. If you track a category and notice that the top five products all experience stockouts during a specific quarter, the entire market is under-supplied. This is your entry point. You can launch a competing product knowing that the established players cannot keep up with the demand.

You calculate the duration of the stockout by counting the days until the rank returns to its previous baseline. If a competitor is out of stock for 30 days, they just handed you 30 days of unfulfilled customer demand.

Filtering social signals for buying intent

Social media engagement is not a demand signal. A million views on a short-form video means the content was entertaining. It does not mean anyone opened their wallet. To find actual demand signals, you must look at what people do after they watch the video and buy the product.

Go to product review sections, industry forums, and developer Q&A sites. Look for the friction points. When a product is trending upward sustainably, the discussions shift from general curiosity to specific technical troubleshooting. If a new type of ergonomic keyboard is just a fad, the comments will be about the aesthetic. If it is a lasting trend, you will see highly technical posts in online communities discussing switch replacements, firmware flashing, and wrist angles. Those detailed troubleshooting posts are your demand signals. They prove people bought the product, used it heavily, and care enough to modify it.

You can categorize these community comments to measure the maturity of the trend. Create a spreadsheet with three columns: pre-purchase questions, post-purchase issues, and modification guides. Count the number of posts in each category over a month. If 90 percent of the posts are pre-purchase questions like asking if the product actually works, the market is still in the hype phase. If you see a rising number of post-purchase issues and modification guides, the product has achieved real penetration. The issues are your roadmap for the second-generation product.

Automating the discovery of demand signals

Finding these signals manually requires reading thousands of forum posts and review threads. You can build scripts to scrape product reviews and run them through a sentiment analysis tool to extract the most common friction points. Look for phrases indicating a missing feature or a broken workflow.

If you are building software or digital products, you can use Unthinkable. It is an iOS app that surfaces business and app ideas directly from real demand signals across the web. It acts as a discovery tool to show you where people are actively trying to solve a problem, saving you the time of scraping industry forums yourself.

Calculating the cost of entry

Finding a trend before it peaks is useless if the unit economics do not work. Once you identify a product with a rising year-two baseline and solid demand signals, calculate your margin before writing a line of code or ordering a prototype.

Take the average retail price of the top three competing products. Subtract your estimated manufacturing or hosting costs. Subtract a 30 percent margin for customer acquisition. If the remaining number cannot sustain your business operations, walk away.

For example, if a physical product sells for $40, your landed cost from the manufacturer is $15, and your acquisition cost is $12, you have $13 in gross profit per unit. If your monthly operating expenses are $2,600, you need to sell 200 units a month just to break even. If the historical search floor we calculated earlier only supports 500 total monthly sales across the entire market, you will never hit that break-even point without capturing 40 percent of the market. That is mathematically improbable for a new entrant.

Avoiding the false positive of a crowded market

A crowded market is not always a bad sign, but a market crowded with identical clones is fatal. When a trend is peaking, you will see dozens of products with the exact same feature set and pricing model. The only way they compete is by lowering the price. This is a race to the bottom.

You want to enter a market where the competitors are solving the problem in fundamentally different ways. If one company is selling a high-end, complex solution for enterprise teams, and another is selling a cheap, bare-bones version for students, there is a gap in the middle. The presence of different approaches means the market is still trying to figure out the optimal solution.

That is the exact moment you want to enter. You can analyze the product reviews for both extremes. The enterprise users will highlight the bloated interface. The student users will highlight the missing features. Your product sits perfectly in that gap.

The discipline of walking away

Most founders fail because they fall in love with a product idea instead of falling in love with a market need. When you use search history floors, average sales ranks, and technical demand signals, you will invalidate 90 percent of your ideas. This is the correct outcome. The goal of product research is not to justify building the thing you already want to build. The goal is to find the one trend that is mathematically proven to be growing steadily.

You do not need to be the first person to discover a trend. Being first usually means educating the market, which is expensive and time-consuming. You want to be the person who enters in year two. Let the early adopters waste their money on viral marketing and customer education. You can sit back, measure the baseline growth, read the troubleshooting threads in online communities, and build the second-generation product that actually solves the problem.

Questions

How long should I track a product's sales rank before entering the market?

You should track the average sales rank over a 90-day period and compare it to the previous 90-day period. This provides enough data to smooth out temporary influencer spikes and weekend anomalies. If the average rank improves across two consecutive quarters, the baseline demand is growing.

Why is the second year of a trend more important than the first?

The first year of a trend is often driven by viral marketing and early adopters who may not represent a sustainable market. The second year reveals the actual baseline demand after the initial hype fades. If the lowest sales month in year two is higher than the lowest month in year one, the product has lasting utility.

How do I know if a drop in sales rank is a trend dying or a stockout?

A dying trend shows a gradual decline in sales rank over several weeks or months. A stockout presents as an immediate, overnight drop from a strong rank to near zero. If multiple top competitors experience these sudden drops simultaneously, the market is severely under-supplied.

What is a healthy decay rate for search volume after a viral spike?

A healthy decay rate is less than 50 percent over the three months following a peak. This indicates the market is retaining a large portion of the new audience and transitioning from early adopters to mainstream buyers. A decay rate of 90 percent or more suggests a temporary fad.

Part of the guide How to Find a Business Idea People Will Pay For, which puts every step in reading order.