Stock Photo Research Tool Rebuilt

Update: Moved to a new host. Send me feedback if you want this tool to stay up and running.

I’ve built a new research tool prototype with new tech, to help you find ideas/topics/subjects for your stock images, i.e. what to shoot. The old one was hitting the database too hard, and built on difficult to maintain technology, so it’s time for an update.

Since I put it live yesterday I’ve had some positive comments about it’s speed and clean looks, but also got some complaints about how it’s no longer possible to copy/paste keywords in bulk from the page, and I wanted to address this particular issue because this is intentional:

Keyword Spamming is BAD

The keywords displayed on the old picNiche tool were never meant to be used in bulk as keywords for stock photos, they are related to the original search made in the tool, but applying them in bulk to photos is equivalent to keyword-spamming (which I am very strongly against) which directly harms the quality of search results presented to image buyers, and harms conversion rates for sales (as buyers then have to spend longer wading through images before they find the ones they want).

Every time your image appears in a search because of a keyword, but is not purchased because that keyword is not relevant, that image is downranked by each agency’s algorithm, and so appears and sells subsequently even fewer times.

For keyword suggestions I highly recommend the stock photo keywording tool provided by MicrostockGroup which provides a more accurate way to select only keywords relevant to you image, and by applying only relevant keywords (not just popular ones) you will improve your sales significantly. Or of course if you don’t want to take a little time to keyword your photos yourself, you can buy keywords from picWorkflow’s highly experienced keywording team 🙂

How to use the research tool to decide what to shoot?

The stock photo research tool gives you an arbitrary rating for possible search queries, (i.e. the ‘conversion rate’ of images matching that search at a popular agency). Enter a search query you think an image buyer would enter, and the ‘rating’ column shows a number predicting the chance of a sale if the buyer used that search (so be mindful of spelling, and use common sense).

If you’re a really awesome photographer, anything over about 20-25 is good; if you’re producing average quality then something over 100-200. Anything over a few thousand is probably a false-positive with either too few images to predict reliably, or unlikely search phrase; though again a little common sense is needed… use search queries a human being might actually search, this tool will not help you ‘game’ the system.

More features will be added

I’ve not finished the new site yet (it’s just a prototype) so will update the page with some more information in future. As suggested by others today I will be re-adding the rating-over-time graph and a shorter set of keyword suggestions to help with quicker checks on related niches/topics/ideas (but I will intentionally make it difficult to copy/paste keywords in bulk to discourage keyword spamming). I will also have it suggest related niches from the search history once a bit more data has been gathered (that I can test suggestion algorithms with).

Please give it a try then let me know what you think, I’ll try and accommodate any (except of course the spamming helper) suggestions or requests 🙂

This entry was posted in Research and tagged , , , , , , , , , , , . Bookmark the permalink. Post a comment or leave a trackback: Trackback URL.


  1. Judith
    Posted September 16, 2014 at 1:33 pm | Permalink

    Nice to see this!!! But bit confused.
    Have been trying out the search and got:
    VF DF Rating
    search 1 EN 11.5 2 23
    search 2 EN 16.389 0.973 15.945
    search 3 EN 5.373 0.583 3.133
    search 4 EN 2.149 0.349 0.749

    Is 23 relatively good (as in there is some likelihood of this one selling) and 0.749 bad (as in don’t waste your time unless it’s really fab)?

    Also – sorry if I’m being dim – but does VF stand for View Frequency and DF for Download Frequency? Or something else….

    Love the idea of a tool like this but feeling outwitted by it at the moment! Hope this is helpful! And best wishes!

    • bobbigmac
      Posted September 16, 2014 at 2:25 pm | Permalink

      V/F = Views per File (suggestive of relative search volume)
      D/F = Downloads per File (suggestive of relative purchase rate)

      It’s hard to answer something like “23 is good” without knowing the search used, something like “business man with handshake in a meeting with businesswoman on a saturday” is ridiculously long so has a high rating, but suggests an idea which can be refined to a simpler idea, and still gain a valuable rating. As here:

      In this example the phrase “relaxing business meeting” shows a much lower end rating than the longer searches, but because it’s a much more natural phrase and simpler idea, it’s more likely to be typed by a photo buyer (and still presents a pretty nice topic for a shoot). In this case I’d consider how to make a meeting look like it is casual/relaxed, or maybe some kind of emergency like “oh no, the website just went down, we’re losing customers, lets have a meeting about it at 10pm on a saturday night”, then that’s the shoot 🙂

      It’s about using your common sense to decide whether the phrase you entered is likely to be searched, the numbers are there to help you decide if that idea has sales potential (i.e. is not already oversupplied) and to refine/find appropriate words and phrases to describe those images.
      The numbers are most useful when you compare them to the numbers for similar searches 🙂 Without you sharing what you searched for it’s impossible for me to judge a specific case, as a general guide anything between 25 and a few thousand means there’s room in the market, but as I said above, if your search is meaningless so are the numbers… or garbage in = garbage out 🙂

      • bobbigmac
        Posted September 16, 2014 at 2:29 pm | Permalink

        I guess short version is, avoid shooting specifically for any reasonable search phrase, where the result is less than 25ish (unless you’re REALLY good) or more than a few thousand (very high ratings suggest there is maybe room for 1 or 2 images, but I wouldn’t arrange a whole shoot on that idea).

  2. Leia
    Posted July 26, 2015 at 1:56 pm | Permalink

    Yep, well, keyword spamming is always bad, both for stock photos as well as SEO.

  3. Andrew
    Posted December 17, 2015 at 3:21 pm | Permalink

    Nice tool, but can be really brilliant if support search in different media types, I.e. Pic, video, vector

    • bobbigmac
      Posted January 29, 2016 at 6:13 pm | Permalink

      It doesn’t really apply, since that would just skew the results to be less accurate, since this tool is assessing the quality of the concept, rather than the type of media. I used to have the old version filterable by media type but the results were wildly inaccurate when restricted to a single type.

Post a Comment

Your email is never published nor shared. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>