Lee has a great post on his blog about the Long Tail and the application of some of its ideas to microstock. The post and comments are well worth a read, but Lee’s summary gets to the heart of the matter.
- If you want to create photos in the most popular subjects, expect that you’ll be competing on quality
- If you want to create photos in long tail niches, keep your costs (and expectations) modest
- Photos with a small market (low demand) might earn more in non-microstock markets
If you want to reap the greatest rewards, you have to find a way to play in the competitive categories around business, lifestyle, medical shoots etc. It’s possible to find areas where competition is relatively low, but volume (and therefore potential) rewards are likely to be lower. To cover many topics in the tail to boost your aggregate earnings, you have to pay attention to your production costs to make sure you can stay profitable.
Ultimately, as with everything, there’s no free lunch and you need to assess the market, your abilities and put in the work.
Environmental themes are common in microstock and I thought it would be helpful to share the top search keywords associated with ‘energy’ to aid in planning and keywording energy related microstock concepts.
Top Search Keywords (via Google Adwords Query Tool)
The big takeaways here are is that ‘solar energy’ is searched for twice as often as ‘wind energy.’ I also initially found ‘jobs’ surprising. With hindsight, it makes current sense given the current economic climate and that concept might be worth exploring as part of a shoot.
Google Insights Trends
I plotted 2008 data to get a sense for full-year seasonality and apart from a steady rise from Jan to April, there doesn’t appear to be a significant seasonality barring a decline from November to Jan which is likely due to Thanksgiving & Christmas.
While drilling into solar energy a little more, I found the regional data interesting as well. The top five regions are Nigeria, Pakistan, India & South Africa from a search volume perspective.
Implications for Microstock
Given the above data, it’s clear that solar should be at the top of your concept list and given the regional trends, varying the ethnicity of your models is probably worthwhile as well.
I came across a great post by Torchdesigns on Dreamstime which talks about what designers are looking for from stock photography. It’s a great post and well worth your time to read but the summary is they are looking for:
- Something specific
- High quality
- Harmony with the other elements
When I have purchased images for use in presentations, the above points really ring true. For example, in the beer vs. wine post, I wanted images that would look good together and I looked at a ton of beer & wine images before choosing the ones I did.
Torchdesign’s final point can be applied in a wide range of scenarios
Of all of the work that I have done It was always important to have the client in mind when considering the finished product. What does the client want the finished piece to express about them?
No matter what you do, there’s always a client. Keeping him or her in mind raises the ultimate quality of your final product.
(As an aside, I think the harmony point could seed a really interesting search filter idea – not just similar images, but images that might look good together. e.g. if you add one shot isolated on black to a lightbox, maybe show more shots of other subjects on black when searching for different subjects. This would allow you to build a cohesive collection of images for use in a subsequent design.)
I analyzed and plotted the latest projections for degrees conferred from the U.S. Department of Education to show gender related trends. The graph below shows historical and projected data by gender for all degree types (Associated, Bachelors, Masters & Doctoral) from 2000-2015.
More Women than Men Graduating. Older Student Population (25+) Growing faster than the 18-24 Segment
Women have been graduating in greater numbers than men and the percentage of women graduates (shown in green) is continuing to rise.
Another interesting nugget was that the growth rates in enrollment for students aged 25 and older is projected to be significantly higher than for students aged 18-24. by 2018, students aged 25 & older are expected to account for 41% of the total college-enrolled population. (Source: US Dept. of Education – Projections of Education Statistics to 2018)
Potential Implications for Microstock Shoot Planning
I’m sure that most of you are already including men and women in your student & graduation shoots. It may also be worth experimenting with using older models in these shoots given the long-term trends around college enrollment.
While playing around with Google Insights data for Valentine’s day trends, I was somewhat surprised to find that there are on average 6 times as many searches for ‘valentine’s gifts for him’ as there are for ‘valentine’s gifts for her.’
I guess men just stick to flowers, chocolates & jewelry while women have a harder time with men’s gifts. On a slightly related note, flowers peak on Valentine’s Day & Mother’s Day with jewelry ruling the roost at Christmas time. I was a little surprised to see such a small uptick for jewelry at Valentine’s day but I suppose it’s fairly soon after Christmas.
When planning your shoots for Valentine’s day, it looks like flowers & chocolates will be your friends.
As part of my analysis of seasonal trends, I looked at LookStat‘s 2008 data for images tagged with ‘Easter’ as a keyword. In addition, I did some quick search keyword analysis to see if there were any useful pieces of data.
2008 Royalties for Images Tagged with ‘Easter’
Easter was on March 23, 2008 in the USA. As you can see from the chart above, the 7-day moving average crosses the daily mean about two months before the holiday itself. In this case, January 5. The peak of the curve is one week before the holiday on March 16. This is very similar to the patterns I saw in the Thanksgiving data.
Easter Eggs are 3 time More Popular Than the Easter Bunny
The Google Trends data for Easter shows an identical pattern to the microstock earnings data shown above. One very interesting nugget that jumped out was in the search keywords. There are 3x as many searches for Easter eggs as there are for the Easter Bunny. I’m not sure why this was surprising to me but i had expected the Bunny to fare a little better.
Conclusion
Easter will be on Sunday, April 4, 2010 in the USA. There’s still time to get your Easter images online before the sales really start to ramp in January/late Feb. Also, don’t forget to decorate and shoot those Easter Eggs!
When analyzing the data for Thanksgiving image sales, I was intrigued by how close to the holiday the sales peak was. I assumed that the early increase would be print sales and that online uses would prevail closer to the time of the holiday. My hypothesis was that if I dug into the earnings per download ($/DL) data, I would see a high value early (print) that declined as online sales grew.
This was not the case. The real peak was the weekend before Thanksgiving.
2008 Earnings per Download for Images Tagged with ‘Thanksgiving’
Earnings Per Download Peak One Week Before Thanksgiving – Not What You’d Expect
The moving average for both earnings & downloads are shown as an underlay on the chart above. As you can see, there is jump in $/DL in August, three months before Thanksgiving, which coincides with the initial volume peak. The $/DL declines a little until mid-September but then begins climbing again and peaks on November 23rd which is the weekend before Thanksgiving.
The early peak at the end of August does seem to coincide with the notion of print sales & magazines. The increasing number however suggests that there are print uses all the way up until the holiday itself. It seems unlikely that individual online buyers would buy large sizes or ELs.
Possible explanations:
- Print lead times have compressed and corporate buyers are making final decisions much closer to the release date than initially assumed
- Retail Sale Circulars & Newspaper inserts etc may play a large role and their volume runs may require ELs. This would fit the notion of the peak being the weekend before Thanksgiving when all the specials etc go out
I’d love to hear your experiences and thoughts on what might be happening. We’ll be examining data for other seasonal events to see what we can find.
One of the things that interests me around seasonal holidays is the timing of changes in sales velocity of microstock images associated with those holidays. The goal is to identify when to shoot & upload images in order to benefit the most from the seasonal spikes in interest.
The chart below shows a 7-day moving average of the royalties for images with the keyword ‘thanksgiving’ in 2008 and it also overlays the daily mean for those images on the same chart.
2008 Royalties for Images Tagged with ‘Thanksgiving’
The royalties crossed the daily average for the year on August 22, peaked on November 23rd and then dropped off from there. The total volume of royalties between August 22 & November 28 accounted for over 63% of the year’s total. You had to be in early to maximize your benefit, but it would have been worthwhile to continue submitting images throughout September & October.
It seems reasonable to see sales start to climb three months before the holiday itself, but it was interesting to see them peak so close to the actual date of Thanksgiving, which was November 27, 2008 in the USA. I’m speculating here but, the early start would seem to coincide with the print & retail calendars and the continued strength suggests more online uses as we got closer to the holiday itself.
(As an aside, it may make sense to buy props and shoot immediately after a holiday when prices are low and your outdoor shots will still look season appropriate. I’m sure some of you out there are already doing exactly this.)
I’d love to hear your thoughts on this if you get a chance. What has your experience been with seasonal images?
According to a June 2009 Gallup Poll of those who drink alcohol, beer is the most popular drink in the US and beats out wine & liquor as you can see in the chart below. In terms of dollar value, rough estimates from the Department of Commerce & Craft Brewer’s association put the wine industry at $30 Billion and the beer industry at $90 billion.
Relative Image Volume of Wine vs. Beer in 2008
The microstock popularity picture is very different. Images of wine outnumber those of beer by a factor of 4:1. There could be a number of factors behind this such as the fact that wine is more aspirational than beer from a marketing standpoint. (I also wonder how much of it is driven by the fact that wine may be more satisfying photograph.)
Royalties Per Image of Wine vs. Beer in 2008
The RPI numbers tell a different story. Images of beer earn 70% more than those of wine on a per-image basis. My hypothesis is that this a result of a supply/demand imbalance, especially when you take into account the overall market and preference numbers.
I’d love to hear people’s thoughts on this and to hear if this lines up with your anecdotal experience.
An interesting aside: according to Google Insights, the top search term related to wine in 2008 was ‘red wine’ and the top term for beer that year was ‘beer pong.’