Heroku Add-On redesign: Impact to Spreedly lead flow

Posted on March 5, 2013 Justin Benson

Spreedly is a a credit card vault in the cloud that connects to more than 45 payment gateways. Vault your cards with us, significantly reduce PCI compliance headaches via a transparent redirect, and work with one or multiple payment gateways over time or simultaneously. We have also added alternative payment types like Dwolla, GoCardless and of course PayPal. (Read about our recent re-launch here)

Heroku’s add-on marketplace seemed like a great potential channel for us. Building an app on Heroku and want to accept payments? Stripe + Dwolla? Braintree + PayPal? Stripe + Braintree + PayPal + Authorize.net? Just select Spreedly as an add on and you’re good to go. Shortly after pushing our new offering live in March 2012 we turned to integrating with Heroku’s add-ons. July 2012 was our first full month in the add-ons store. The following chart shows trial sign ups from Heroku, from all other sources (Spreedly) and then the combined total.

Heroku July

So around 28% of all trial sign ups were coming from Heroku in the month of July. We had 3 customer begin on payment plans too – albeit our smallest most basic one. Still, any time you can add a new channel and increase your lead flow nearly 30% you’re happy!

September is a great month in the Northern Hemisphere. Everyone’s back from summer vacation and starting up projects. Smaller projects you put off over the holidays or enterprise projects you have in mind for the next calendar/budget year. We were still a bootstrapping service and relying on word of mouth but we were happy do 71 trial sign ups in September. It was our best month to date. But we were struggling with Heroku. Heroku Sept

We’d gone from 28% of leads attributable to Heroku in July to just 14% in September. It seems we’d fallen into the “We were hot while we were the new kid on the block” and by September interest had cooled off. Worse, we didn’t add any new paying customers in the month of September. We were really happy with overall traction and interest in Spreedly – but Heroku was disappointing.

A blip or a trend?

November and December are both a bit problematic in the US. Thanksgiving is November in the US and then there is Christmas. So that can impact data. Still, here’s November (which did predictably drop off in the last week for all new activity)

Heroku Nov


Again, right at 15% of sign ups coming via Heroku’s add-ons. We’re now pretty sure we’re staying at this level for the foreseeable future.

Heroku then completely redesigned their add-on store.  We were pushed live during the first few days of January 2013. (We needed to make changes to our profile to support the new design so we didn’t go live as soon as the new format went live) The response was immediate and impressive.

Heroku January


Heroku delivered 3 times as many leads to us in January as they did in November. That’s a big bump! In a relative sense they were also back to being 30% of all leads for us again.

I wanted to wait and see what happened with February. Perhaps it was just a blip again? Well through the first 3 weeks of February we were still on track at 30%

“Coffee’s for closers”

i) Leads aren’t sales. We did start adding new customers again in January – which really hadn’t happened much in Q4. It’s too early to tell if Heroku leads are worse/same/better quality than other sources.

ii) Pricing problems: Heroku’s add-ons don’t support metered pricing. So we have to create plans that sort of look like our pricing. However, they don’t map exactly and that matters to some customers. We’re pretty sure that some Heroku users prefer to sign up directly vs activate purely via Heroku. Sort of like “Shop in Orbitz and buy direct from the airline website” mentality (sorry Heroku if you’re reading this!)

iii) This last point makes us less concerned about the revenue that flows through to directly from Heroku. Again, while starting up, the exposure itself is fantastic.

iv) We’re the only option under “Billing” right now in Heroku. I’m sure that will change over time (after we publish this?) That will impact interest once it happens no doubt.

With our major re-launch and announcing funding at the start of March our March metrics will break from the previous trend. So this is a good point to take this snapshot and lay it to rest. In summation Heroku’s add-ons have been:

  • Great free marketing for a bootstrapped startup via the exposure
  • Ok on new customer wins
  • More weighed to “tire kickers” than folks who sign up on our site directly

Of course, our results will be directly impacted by how relevant we are to the Heroku audience and profile. We tend to find our best prospective customers are running larger, complex commerce/payment platforms.

Any comments might be here.

Sign up for blog post updates

Get notified about new Spreedly articles on PCI compliance, gateways, & more.