LFGSS in numbers

More numbers, because I was curious.

Totals:

• 5 million comments
  
• 152 thousand conversations
  
• 400 events
  
• 49 thousand registered members
  
• 24 thousand active members (logged in within the last 30 days)
  
• 2.3 million pages indexed by Google
  
  

Each month (avg):

• 6.8 million pages viewed (if comparing to old vBulletin numbers, this site has fewer interstitials on redirects and page views is approximately triple the old numbers in the first post of this conversation)
  
• 1 million visits (sessions) from about about 175 thousand users (better measurement than it used to be, unique cookies over visits, rather than calculated approximation)
  
• £550 costs (similar to first post costs, but 3 x traffic and users served)
  
  

Each day:

• 33 thousand unique visitors
  
• More than 1,000 concurrent people online
  
• 3,000 new comments
  
• 80 new conversations
  
• 35 new members
  
• £20 in donations
  
• 120 thousand pages indexed by Google
  
• 30 thousand search results shown on Google
  
• 6 thousand click-throughs from Google
  
  

Some of these numbers come from Google Analytics, which doesn't have complete info (ad-blockers, natch), but additionally:

• 75% of traffic comes from desktop browsers
  
• 22% from mobile phones
  
• 3% from other stuff (tablets, games consoles, TVs, whatever)
  
  

On the money stuff, affiliates has almost dried up, donations support us exclusively.

There is an impending issue on the money... our costs will rise in the next year.

We use lots of free things or cheap things and these have thresholds for how much you can use. We use a lot, we're about to cross over thresholds on things like Google Maps (so I may change the default map provider, but even then I'd have to pay a few hundred £ per month to get the next cheapest based on Open Street Maps). Same is true of even Google Analytics which I use to get info on devices that use the site, performance of the site for people overseas, etc... I can't afford the paid version so on this we'll start doing client side sampling to reduce the data we send.

All in all... very busy site, the numbers remain huge and still growing, it costs us as much today as it did 6 years ago but we serve 3 x the number of traffic.

It's all good, but eyes to the future as I need to fix monetisation to ensure that it doesn't crunch when costs rise.

Why did affiliate revenue disappear?

Is it driven by the lack of guest traffic as was the case two years ago?

Some companies exited affiliates altogether:

i.e. Rapha, and British Cycling (BC do partnerships with companies like Evans, but few to no affiliates with forums, etc)

Some companies revised their agreements:

i.e. eBay pay far less on bids, have a far shorter cookie which means unless someone bids immediately we aren't credited with it, and they also reduced the % paid out

Some companies geo-fenced their payments:

i.e. As a UK Amazon account I no longer get the US Amazon referrals

Some companies just reduced their payments:

i.e. Wiggle and Chain Reaction merged and normalised their payments to the lowest % either was using.

Some companies added delays to handle returned items:

i.e. If you earn a referral, you have to wait 30-60 days for it to be confirmed, I don't believe as many people return items as transactions later are not confirmed but as a publisher have no power to dispute this.

Some users use browser add-ons:

i.e. cashback extensions that hijack the cookies so we are never even acknowledged as the referrer

Some users use cashback codes:

i.e. agreements state that if such codes are used the payments are lower or non-existant

In essence: From almost every angle, the affiliate revenue has been sliced down to the bone.

There is still revenue, but like when eBay did their change and that caused the biggest collapse in revenue, now everyone else has done changes at much the same time and that's wiped out the lion's share of the rest.

What remains isn't a reliable source of revenue, nor predictable. It can still be a couple of hundred £ in a month, but then will sometimes surprise me by being £30 (and the finances fall short that month) and then a couple of months later spike to closer to £500 (and then I leave the money in PayPal and it irons out any drops in the following months).

The unpredictability is really the issue. Our costs are stable, the donations are fairly stable (gently declining due to me not really pushing it and reminding people, but made up by a few exceptional one-offs each month in the £50 ballpark).

Stability is ultimately the goal on revenue, to be comfortable not thinking about it, not chasing it, it just ticking over peacefully and letting us all get on with stuff.

Presumably at this point you're pretty glad that Microcosm wound up?
(as affiliates was the main aim for revenue source?)

Yes.

It definitely is a risk for LFGSS, but even with funding this would have forced Microcosm to push hard for revenue from the classifieds, events, and other sources.

I may have to do that eventually, but I can do it right... gently, and with respect to finding a balance in it all.

Microcosm would have put everything at risk, now that affiliates are collapsing as a business model. The idea of Microcosm was that scale insulates from that, but the scale needed with affiliates as they are is far greater than what I had originally projected.

And really cool is that I moved my Amazon S3 bucket... don't ask why... and I got to see numbers of how much storage space is really used by how many things.

Attachments and Avatars

• 92GB on disk
  
• 200 thousand items
  
  

It's massive.

And ignoring the fact that this month I had to pay data transfer costs, it costs us virtually nothing to store and serve this most months... lower than $20 per month.

Amazon S3 is incredible.

...and very few community / forums achieve scale...right?

Even if you have a community of communities

A lot do.

It's just time.

If they survive the first few years, they will grow organically constantly.

The community of communities is still a sound idea, the compounded organic growth of each community does create significant growth.

The flaw for ventures walking this path, is that the time it takes for the compounding effect to happen for new sites is too long for the life of the venture. Especially given that large communities tend not to migrate to different tech... preferring to stay with whatever is working for them now, as the effort of moving is seen as worse than stagnation.

Can you do these numbers again for now please? IS the site still growing? Have we reached peak fixi skidder wanker yet?

fix monetisation

Have you managed to do this?

I strongly suspect that the platform is growing, but that LFGSS is at peak. I'll run the numbers this weekend and will see.

Totals:

5.5m comments
170k conversations
588 events
55k registered members
17k active members (logged in within the last 30 days)
645k pages indexed by Google (I removed duplicate comment pages via robots.txt)

Each month (avg):

7.1m pages viewed (if comparing to old vBulletin numbers, this site has fewer interstitials on redirects and page views is approximately triple the old numbers in the first post of this conversation)
890k visits (sessions) from about about 145k users (better measurement than it used to be, unique cookies over visits, rather than calculated approximation)
£740 costs

Each day:

22k unique visitors
More than 1,100 concurrent people online
2,700 new comments
60 new conversations
35 new members
£18 in donations
50k pages indexed by Google
18k search results shown on Google
4k click-throughs from Google

Some of these numbers come from Google Analytics, which doesn't have complete info (ad-blockers, natch), but additionally:

71% of traffic comes from desktop browsers
29% from mobile phones
0% from other stuff (tablets, games consoles, TVs, whatever)

Overall from this... costs have risen (predominantly due to currency exchange rate against USD and EUR not being favourable and most services are paid in USD or EUR), donations have declined (only fractionally, but it's still break even so it's fine), the core users are using the site more (page views) but there's probably fewer new users participating so average daily comment count has dropped slightly, there are still new people joining at the same rate, but the number of guest visitors from Google has dropped.

LFGSS has probably peaked. It probably was earlier this Summer. That's OK too... was always going to happen one day :)

Oh... another number... emails.

We used to send, about a year ago, around 50k emails per month in notifications.

We now send 79k per month.

Hence my reading that the core group is more engaged... notification emails are only sent again when people visit and the number wouldn't rise significantly if it were more people less engaged... so this is the same or fewer people far more engaged.

71% of traffic comes from desktop browsers

Is this higher than the majority of websites?

I can find out tomorrow 🙂

peaked

It's a free wheeling down hill rollercoaster from here.

I'm gonna hang on until there are only a hundred people left. The good old days.

I remember when it was all fields around here ...

Can we speed it up and have a cull of everyone who signed up during 2008 and afterwards?

What is the split between chat pages and classifieds?

For members, views, etc..

I often feel like this place is a free eBay but wonder if that's just a bias.

Interesting, my perception over the past few years the classifieds is less busy than it was and that non-bike related discussion has increased.

I don't have a split based on which forum, but could probably extrapolate something from the number of comments in forums last year and this year.

My hunch is that bike related chatter via Current Projects has increased, non-bike related chatter has increased, and that Classifieds has slightly decreased (is there a point in which we have all the bikes we want?).

Hey!

So that's Hippy's stats, what about the rest of the site?

is there a point in which we have all the bikes we want?

No

Dam new commer