5 ms·

Revenue & Expenses 2 Years After Starting My Web Business: UniversityTutor.com

30 points by barmstrong 17y ago

patio11 17y ago
I'm broadly supportive of things like this, but have some constructive criticism. 1) Don't graph dissimilar things on the same graph. If you absolutely must, to make a point of how they correlate with each other, do not re-use the same axis. There is no reason Revenue ($) should ever reuse an axis with Contacts (number of). 2) Don't graph anything you can't get actionable data out of. For example, revenue per day in a subscription business is going to be dominated by the anniversary dates for existing accounts. Is that useful to know? No, that is noise. Condense it into revenue per month and then you have a useful top-level overview of whether your efforts are bearing fruit or not. 3) One line on one graph. At most, two. Anything more turns into a solid wall of noise. (You may find it occasionally useful to have a solid wall of noise in presentations to other people. Don't put it on your dashboard! Then you're only fooling yourself!) Separate strategic commentary: Mass profusion of subdomains is not a good idea for SEO purposes. (I assume this happened because it was in fashion two years ago. This is one of the dangers of chasing the new hotness in SEO techniques... though I don't remember that one ever being a good idea at less than black-hat scales.) I'd put the blog on the same domain and, if not interested in maintaining it, hire somebody to continue writing content for me. (I need to take my own advice here, come to think of it, as the BCC-specific blog doesn't get content updates nearly frequently enough.) Start A/B testing. (You knew I was going to say that, right?) One obvious candidate would be adding a featured flag to the subscribing tutors, with the goal of seeing whether it increased a) contacts to them and b) subscription rates. (As soon as you have an answer to Question A it becomes a marketing point: people who pay us $10 a month get 63% more contacts and more than $200 of extra revenue! Dude, its a total no brainer!)
- barmstrong 17y ago
  Great comments thanks for posting this. You bring up some good points on the graphs. This was mostly quick and dirty stuff I hacked together and could certainly be improved. The 'noise' doesn't seem to bother me though and helps me compare them visually, could be a personal preference thing. On the SEO use of subdomains, do you have any links I could read more on that? Anecdotally, my experience seems to be otherwise but I could be wrong. Agreed that the blog needs better maintaining. I'm more focused on my personal blog to do much with it. I've tried getting some tutors to contribute with mixed results. I haven't had much luck with the quality of articles when hiring someone to write, so not sure of the ROI on that. If anyone has ideas would love to hear it. On that A/B test, funny you mention that because I tried almost that exact thing you suggested when 37Signals came out with Haystack.com (now Sortfolio.com), where I made the free tutor account permanent and the paid account got you top placement in search results, highlighted listing, and a 'PRO' badge next to your name. I didn't run it long enough to have 100% confidence but the conversion rate dropped compared to what I have now (first 3 job requests are free, then you must upgrade if you want to continue). I still feel like the biz model could be improved though. If you have other ideas on this would love to hear it. Thanks again! Brian
uvince 17y ago
It's great to see posts like this to help us remember: not only do you have to build it but you have to run it. Running it and managing it to ultimate success may very well be the hard part.

30 points by barmstrong 17y ago

I'm broadly supportive of things like this, but have some constructive criticism. 1) Don't graph dissimilar things on the same graph. If you absolutely must, to make a point of how they correlate with each other, do not re-use the same axis. There is no reason Revenue ($) should ever reuse an axis with Contacts (number of). 2) Don't graph anything you can't get actionable data out of. For example, revenue per day in a subscription business is going to be dominated by the anniversary dates for existing accounts. Is that useful to know? No, that is noise. Condense it into revenue per month and then you have a useful top-level overview of whether your efforts are bearing fruit or not. 3) One line on one graph. At most, two. Anything more turns into a solid wall of noise. (You may find it occasionally useful to have a solid wall of noise in presentations to other people. Don't put it on your dashboard! Then you're only fooling yourself!) Separate strategic commentary: Mass profusion of subdomains is not a good idea for SEO purposes. (I assume this happened because it was in fashion two years ago. This is one of the dangers of chasing the new hotness in SEO techniques... though I don't remember that one ever being a good idea at less than black-hat scales.) I'd put the blog on the same domain and, if not interested in maintaining it, hire somebody to continue writing content for me. (I need to take my own advice here, come to think of it, as the BCC-specific blog doesn't get content updates nearly frequently enough.) Start A/B testing. (You knew I was going to say that, right?) One obvious candidate would be adding a featured flag to the subscribing tutors, with the goal of seeing whether it increased a) contacts to them and b) subscription rates. (As soon as you have an answer to Question A it becomes a marketing point: people who pay us $10 a month get 63% more contacts and more than $200 of extra revenue! Dude, its a total no brainer!)

barmstrong 17y ago
Great comments thanks for posting this. You bring up some good points on the graphs. This was mostly quick and dirty stuff I hacked together and could certainly be improved. The 'noise' doesn't seem to bother me though and helps me compare them visually, could be a personal preference thing. On the SEO use of subdomains, do you have any links I could read more on that? Anecdotally, my experience seems to be otherwise but I could be wrong. Agreed that the blog needs better maintaining. I'm more focused on my personal blog to do much with it. I've tried getting some tutors to contribute with mixed results. I haven't had much luck with the quality of articles when hiring someone to write, so not sure of the ROI on that. If anyone has ideas would love to hear it. On that A/B test, funny you mention that because I tried almost that exact thing you suggested when 37Signals came out with Haystack.com (now Sortfolio.com), where I made the free tutor account permanent and the paid account got you top placement in search results, highlighted listing, and a 'PRO' badge next to your name. I didn't run it long enough to have 100% confidence but the conversion rate dropped compared to what I have now (first 3 job requests are free, then you must upgrade if you want to continue). I still feel like the biz model could be improved though. If you have other ideas on this would love to hear it. Thanks again! Brian

uvince 17y ago

It's great to see posts like this to help us remember: not only do you have to build it but you have to run it. Running it and managing it to ultimate success may very well be the hard part.