How Mergeflow extracts market data from texts

For any query, Mergeflow shows among its results also market estimates. This includes market segments, size, and growth estimates. Mergeflow visualizes market data in interactive charts.

Where Mergeflow gets its market data

Mergeflow has an algorithm that checks every incoming piece of content for market estimates. This includes news, blogs, press releases, websites, and other sources.

Importantly, Mergeflow does not calculate market estimates. It only extracts these data from web-available contents.

The data in Mergeflow's Markets view

In addition to the name of the market segment, market estimates usually have three components:

  • Estimated size
  • Estimated growth rate
  • A time frame for which the growth rate is estimated

Different sources use different formats for conveying this information. Some sources report a size estimate for a certain point in time plus a growth estimate; others report two size estimates for two points in time (from this the growth rate can be inferred); and different sources use different points in time.

Here is an example of how Mergeflow displays extracted data (the query here was "smart city"):

The columns:

Type: "Direct" means that the market estimate directly matches your query, or parts of your query. In this case, the query was "smart city" OR "smart cities". "Context" means that the market estimate came from a document that matched your query, but the market segment itself did not. In many cases, these "context" estimates are particularly useful because they provide you with context. For instance, for "smart city", "fiber optics" or "5G infrastructure" are constituent technologies.

Market segment: This is the name of the market. Sometimes market segments are qualified geographically (e.g. "US fiber optics market" vs. "Asia fiber optics market" vs. "Global fiber optics market"). Not having a geographical qualifier does not necessarily imply "global". It simply means that geography was not specified in the estimate.

CAGR (%): If a market estimate reports a CAGR, this is what is reported here. If a market estimate does not report a CAGR but reports two market sizes at two points in time, Mergeflow does a linear extrapolation to infer the CAGR from the two size data points. Notice that in reality, CAGR may not be linear; it may change its slope over time. But since we do not have this information available, we do the simplest thing possible, which is a linear extrapolation.

2017 (mio. USD): This is the size estimate for a market, in Mio USD, for "three years back from current year" (it is 2020 now, so 2017). If a market estimate is available in a different currency, we convert the currency into USD at the daily exchange rate (01 January 2017 in this case). If a market estimate reports a size estimate for a different point in time, we take the size estimate for this point in time and use the CAGR estimate to linearly estimate the "three years back" size.

2020 (mio. USD): This is the Mio USD size estimate for "current year" (now = 2020 here). Just like for the "three years back" estimate, the "current year" estimate is reported directly if it is available in the source. If not, we again take size estimate at a different point in time and use the CAGR to to a linear estimation for the "current year" size.

2025 (mio. USD): This is the Mio USD size estimate for "five years from now" (now = 2020 here). This figure is calculated in the same way as the other size estimates, in cases where the number is not explicitly available from the source.

Report Date: This is the date at which the market estimate was reported.

Show details: Clicking here lets you access the document(s) from which the market estimate was extracted.

Why Mergeflow sometimes reports more than one estimate for the same market segment

From the same query as above, "smart city", here are three estimates for the same segment, "cellular IoT":

While all these estimates are for the same segment, they differ in numbers, and they are from different dates. So they are not duplicates (which we would de-duplicate and only report once). Because of these differences, we report all of them.

Still need help? Contact Us Contact Us