Found the email. The wonders of gmail
So comparing the ThemeForest sales 1 week before and 1 week after the change we observed that the average sales price had gone up $3.80-ish and sales overall were up 8%. So overall I was pretty happy with the result.
Interestingly before this price change I spent a packet on Google AdWords trying to determine what the optimal sale price point was just by changing ad copy. Was a complete failure as the CTR was the same regardless of what price I put in the copy. Sad really – I thought it was going to be an awesome way to test buyer demand.
Hello – Travis just pointed me this way. Holy toledo this is a long thread now.
We did some analysis of the roughly price rise after the first couple of weeks. From memory it showed a slight slump for the first several days and then picked up overall there after and was positive. Otherwise we would have seriously looked at reverting/changing yet again of course.
I’ll try dig up the summary from one my emails.
I’ve also just asked Sveta, our new analyst (yet to be introduced I think), to have another look at the data now that some time has gone by. We’ll post up some details once she’s had some time to look at it. She’s kind of busy also trying to hire help and a DBA for a data warehousing project we’re working on. With all the millions of members, items and transactions we can no longer easily run complex reports off our operational databases
Overall I’d say its a good problem to have
Ok, so compete is incorrect (or INCORRECT as you shout) and alexa is better (because its picture is sweeter) but still… Would be really nice of you if you could take some more time and answer my question which is beyond arguing meters’ accuracy (I am the thread author after all). Thank you.
Hi IconCreme – Sorry, I didn’t focus that part of the question. Overall sales continue to grow, as do all our metrics. So:
“If you have personally observed a change in trends then this indicates this is local affect, possible due to changing buyer habits or buyer interest than anything else. Now THAT is something worth thinking about – what’s on the mind of our buyers.”
I’m the GM for Marketplaces jumping in here.
First: The graph taken from http://siteanalytics.compete.com/ is completely and utterly INCORRECT .
I’m not sure how they collect their data but it has NO correlation to actual analytics data. Not only are the values completely off, but so is the trend and pattern. They are not even in the same ball park. Might as well use a random number generator.
For comparison you can check http://www.alexa.com/siteinfo/graphicriver.net which shows we are growing consistently. Now I don’t have a lot of faith in Alexa either – these sites are relying on 3rd hand information which is not accurate and is often completely wrong.
So how can both Compete and Alexa show completely different graphs and both be correct? The answer is that the data is not reliable.
Our actual analytics data shows traffic continues to grow, month on month since [Edit: GraphicRiver] launch in Dec 2009. Similarly deposits, sales, sign ups all shows continuous growth month on month since launch. Rest assured, if they weren’t then I’d be the first one to note a change and I would be the first one worrying.
If you have personally observed a change in trends then this indicates this is local affect, possible due to changing buyer habits or buyer interest than anything else. Now THAT is something worth thinking about – what’s on the mind of our buyers.
WOW ! 2 Million Members – Totally awesome achievement everyone!
I can’t believe I missed this by an hour wasted gardening (yes I have a mundane life rather than watching the moment tick by.
I think Julian mentioned that the number also includes disabled accounts so we’ll be looking to see how we can correct this in the footer. Maybe we will get to see the 2 mm mark twice.
Not absolutely important to get an answer, but I’m curios, what the problem was?
We’re still not sure and need to investigate with the bank manager why only some transactions got properly authorised in the system. The main issue for today was to get the payments out – which we did. We also added a step in future to double check the transactions the following day on the online bank statements.
We will however be trying to work out what caused the problem in the first place – redoing work is never fun. Particularly when you have a bunch of people who are grumpy with how the first lot didn’t get done properly.
Credit to you all for not being grumpy – you would have been in completely in your place to create a real storm about this. Thanks for your positive support during the last couple of days – I really appreciated it.