Levelling the Land - The Impact of St. Level's Day

[paulmaplesden]

I’m seldom happier than when messing around in Google Sheets, and I wanted to do some analysis on the impact of St. Level’s Day on seller levels and gigs.

With this in mind, I captured the number and percentage of gigs by level across 16 popular categories in the Fiverr marketplace - two under each of the eight top-level categories. I measured them on two dates, Friday 12 January and this morning, Tuesday 16 January - before and after the St. Level’s Day Massacre.

I have shown the “before” and “after” results below, together with the difference between the two days. A few points to bear in mind as you read through the tables:

• This only measures the total number of gigs by seller level in each category. It doesn’t measure number of sellers - however, number of gigs should be an OK proxy.
• This is publicly available information that I took simply by visiting each category, filtering by level of seller, and copying the results.
• Yellow highlights are purely there as a visual aid - I will explain this a little more in the notes under each table.

The Day the Earth Stood Still (Baseline) Table - Friday 12 January

The yellow highlighting on this table simply shows what percentage of gigs by level are above the average for that level of gigs as a whole.

Armageddon Table - Tuesday 16 January

The highlighting on this table shows us where the biggest changes are. I did this by using the average for the previous table and applying it here, so it shows the biggest shifts in numbers.

Deep Impact Table - Changes Between the Two

Finally, this shows where the changes are. Again, the yellow highlighting shows the areas that moved beyond the average for all gigs of that level. Note that the percentage numbers here are absolute not relative.

This means that if 10% of the gigs were at a certain level before, and only 5% are now, I will have shown a difference of -5% (an absolute value) not -50% (a relative one). This is probably only a definition that matters to stats geeks like me, so you can safely ignore it!

So, what does this tell us? Here are my initial thoughts:

• The overall number of gigs between the two days went down, but only by a tiny amount - c. 1.1K, or less than -0.5%, so this has not (yet) caused many gigs to be removed altogether.
• 84% of all gigs are now in the “New Seller” or “L1 Seller” level, compared to 69% before the change.
• The “video” and “marketing” categories saw the biggest overall percentage shifts away from L2 / TRS.
• Voiceover lost the most TRS, followed by Promo Videos and Social Media Marketing.
• Just over 15% of gigs lost their L2 or TRS rating and were demoted to a lower level - that’s less impact than I was expecting.
• Sellers in the Programming, Writing, and Design categories should really work on getting to TRS, as there’s a lower percentage of sellers and gigs of that level.

I hope this is useful, and I would be fascinated with your thoughts. Let’s hear them!

Note: My maths is not impervious - if there are any errors in my insights or calculations, please point them out to me in a private message, and I will correct them. Additonally, this is not a thread to complain about losing your level, there are plenty of other places you can do that. Thanks!

Anyone who wants to dig into the data a little more and do their own analysis, I have provided access to the Google Sheet here. You will need to “Make a Copy” to get editing rights, then feel free to develop formulaes to your heart’s content.

[matevzs1]

Nice, I love such stats! I will now take some time to check it out in detail. 🙂

It would also be interesting to see what is “usual” baseline, so how many gigs are removed from each category on normal 3 days period. 😉 Maybe I will find some time to check that for few days as 1.1k seems really small number for me, that I think this may be pretty normal in such big marketplace.

Also, for accurate calculation, we should also take into account that everyone who lost TRS badge is now LVL 2 (everyone who lost LVL 2 is now LVL 1, …), so actually a little bit more sellers and their gigs have lost their level.

[ollydave]

I think there should be another snapshot on 17th or 18th of January.

Since it will take another 48 hours to grant the level 2 sellers that are eligible for TRS their badge.

This will further change the shifts/results

[paulmaplesden]

Nice, I love such stats! I will now take some time to check it out in detail. 🙂

It would also be interesting to see what is “usual” baseline, so how many gigs are removed from each category on normal 3 days period. 😉 Maybe I will find some time to check that for few days as 1.1k seems really small number for me, that I think this may be pretty normal in such big marketplace.

Also, for accurate calculation, we should also take into account that everyone who lost TRS badge is now LVL 2 (everyone who lost LVL 2 is now LVL 1, …), so actually a little bit more sellers and their gigs have lost their level.

Anyone who wants to dig into the data a little more and do their own analysis, I have provided access to the Google Sheet here. You will need to “Make a Copy” to get editing rights, then feel free to develop formulaes to your heart’s content!

[paulmaplesden]

Nice, I love such stats! I will now take some time to check it out in detail. 🙂

It would also be interesting to see what is “usual” baseline, so how many gigs are removed from each category on normal 3 days period. 😉 Maybe I will find some time to check that for few days as 1.1k seems really small number for me, that I think this may be pretty normal in such big marketplace.

Also, for accurate calculation, we should also take into account that everyone who lost TRS badge is now LVL 2 (everyone who lost LVL 2 is now LVL 1, …), so actually a little bit more sellers and their gigs have lost their level.

Also, for accurate calculation, we should also take into account that everyone who lost TRS badge is now LVL 2 (everyone who lost LVL 2 is now LVL 1, …), so actually a little bit more sellers and their gigs have lost their level.

If you have the time to do that calculation, and can get to the data, I’d love it if you could share it!

[paulmaplesden]

I think there should be another snapshot on 17th or 18th of January.

Since it will take another 48 hours to grant the level 2 sellers that are eligible for TRS their badge.

This will further change the shifts/results

I think there should be another snapshot on 17th or 18th of January.

Since it will take another 48 hours to grant the level 2 sellers that her eligible for TRS their badge.

I am planning to update this after each evaluation for the next couple of months, so we can see shifts over time.

[matevzs1]

If my calculation is correct, this is the number of levels lost by category:
(If we assume that 1 seller has 1 gig. As this is obviously not true, real data is slightly different.)

[paulmaplesden]

If my calculation is correct, this is the number of levels lost by category:

(If we assume that 1 seller has 1 gig. As this is obviously not true, real data is slightly different.)

If my calculation is correct, this is the number of levels lost by category:

(If we assume that 1 seller has 1 gig. As this is obviously not true, real data is slightly different.)

Awesome! Thanks for taking the time to put this together.