A plain-English look at where our numbers come from, what each metric means, and what it doesn't.

SaleHoo Insights is built to help you judge four things in a consistent way: how demand is moving, whether timing follows a repeating seasonal pattern, what price band similar listings sit in, and whether we can surface relevant suppliers for a product idea. It's a decision-support tool. Not a guarantee engine.

The short version: we combine search-interest history, marketplace pricing signals, and supplier-directory matching into one product view. Demand and pricing run through a consistent U.S. market lens so products can be compared on the same basis. Supplier matching uses your product term to search the SaleHoo directory and return relevant results. The aim isn't to tell you exactly how many units you'll sell or exactly what you'll make. It's to let you compare products with more confidence by putting directional demand signals, recurring timing patterns, typical pricing context, and sourcing options side by side.

Here's how each piece actually works.

How We Measure Growth

Growth is the change in relative customer interest over time. We anchor that interest to a five-year trend series, then cut shorter trend windows from the same underlying history: 6-month, 12-month, and 24-month views.

One thing worth being clear about, because it's the part most people get wrong about trend data. We don't treat growth as a simple first-point-versus-last-point jump. We fit a trend across the selected window, so the reading reflects the broader direction of the series rather than two cherry-picked dates. When the recent stretch is clearly moving faster than the older part, that recent movement can pull the final reading, but only when the data gives us enough support to justify it. This keeps one-off spikes, a freak weak starting month, or a short-lived swing from making a product look stronger or weaker than it really is.

And it's a relative signal, not an absolute sales counter. Values are normalized on a 0–100 scale against the peak in the queried range, so a strong positive growth reading means interest has risen meaningfully inside the measured window. A flat or negative reading means interest held steady or softened. It is not a claim about exact search volume, exact orders, or exact revenue.

Spotting a Real Seasonal Pattern

Seasonality answers a narrow question: does this product show a repeatable month-of-year pattern, or did it just have one good run once? To work that out, we derive monthly interest from the broader demand history and look for recurring peaks and troughs across multiple months and years.

A single strong holiday doesn't earn a seasonal label. Neither does one temporary spike or one eye-catching chart. A firmer seasonal classification needs repeated evidence. The current method favors products with enough observed monthly support and gives more weight to patterns that repeat across complete years than to scattered, isolated months. We're looking for a recognizable rhythm, not a single burst of attention.

So when a product is flagged as seasonal, the data has cleared an internal threshold for a recurring pattern. If there's some directional timing signal but not enough for a firm claim, we'd rather stay cautious than overstate it.

Where the Price Range Comes From

Price range is a market-context signal built from marketplace pricing data for matching listings. In the current pipeline, those price points come from Amazon marketplace search results collected for the product term.

We gather the available prices, then cut the influence of extreme outliers by focusing on the middle 50% of the observed distribution, the interquartile range, rather than the full raw spread. In plain terms, the displayed band is meant to reflect a more typical selling zone, not the single cheapest or most expensive listing the system happened to catch. That makes it more useful for benchmarking and setting expectations on average selling price (ASP).

Read it as guidance, though, not a promise. It isn't your landed cost. It isn't your margin, your ad-adjusted profitability, or the exact price you'll be able to charge. It's a structured snapshot of the pricing environment around similar listings at the time the data was collected.

How Suppliers Get Matched

Supplier matching is the bridge from research into sourcing. When supplier data is available, Insights uses the product term as a search input against the SaleHoo supplier directory and surfaces suppliers whose profiles look relevant.

That's a faster start than a blank search box. It can quickly hand you supplier names, supplier types, countries, brand references, and profile details worth having before you reach out or compare options.

But it's a relevance match, not a guarantee. A matched supplier isn't confirming current stock, exact product identity, minimum order quantity (MOQ), shipping capability, quality standards, or pricing fit for your business. Treat it as a curated shortlist for the next stage of vetting, not the final sourcing decision.

What the Data Won't Tell You

Insights is built to sharpen your decisions, not replace them. It can help you compare products more intelligently. It does not guarantee any of the following:

• Future sales performance or future growth.
• Exact search volume, exact unit sales, or exact revenue.
• Profitability after product cost, shipping, fees, ads, duties, taxes, and returns.
• That a product is legally safe to sell in every market, or free from IP risk.
• That a listed supplier currently carries the exact item you want, at the quality, lead time, or price you need.
• That every product will have enough reliable signal to generate every metric.

That's why it's best read as a research layer: something to narrow the field, spot stronger opportunities, and ask better questions before you commit time, money, or ad spend.

How Often Data Refreshes

Insights data refreshes on a rolling basis rather than rebuilding every product at once. The application reviews products for refresh every hour and prioritizes them by business value and data freshness. Products tied to published reports, actively tracked products, and products showing favorable trend signals move toward the front of the queue.

By default, a product becomes eligible for another full automated refresh once its underlying trend data is considered stale, which in the current setup is roughly a 28-day threshold. Higher-priority products may be revisited sooner. Others may wait longer, depending on queue demand and whether fresh external data is actually available when the update runs.

Some layers move faster than the full product cycle. Supplier matches are fetched separately and cached for a shorter window, and some page-level chart data is cached briefly too. So a small gap between a backend refresh and what you see on the page is normal.

Why Some Products Show Limited Data

Some products are just easier to measure than others. Missing or limited data usually means the available signal was too thin, too inconsistent, too ambiguous, or too incomplete to support a metric with enough confidence.

That happens with products that are very new, very niche, searched under a dozen different names, or bundled under a broad term that mixes several buyer intents. It also happens when marketplace results don't return enough usable price points, or when a product hasn't built up enough month-by-month history for a seasonal read.

Sometimes the system tries a product more than once and still gets no reliable demand signal from the source data. When that happens, we back off rather than keep publishing weak numbers. That's deliberate. If the data doesn't justify a confident metric, showing less beats showing a number that looks precise but isn't.

As more demand history, marketplace coverage, or supplier information becomes available, some gaps close over time. When data stays limited, the safest read is simple: the product currently has a weaker measurable signal, not that the system is hiding a stronger one.

---

Methodology last reviewed: June 2026. Figures such as the ~28-day refresh threshold and the hourly review cycle reflect the current pipeline and may change as the system evolves.