Plumping MRR from $1000 to $4000 USD

Monthly Recurring Revenue (MRR) is the most important metric when running a SaaS company. You want your MRR to be as plump as possible. On February 13th MRR was at $1000. Last week MRR exceeded $4000.

What happened across the past seven months? Why did MRR increase so much in the last week?

Letting Time Pass

A big part of it was just letting time pass and steadily getting more users. Looking at the graph there’s only one spike. That being said, in the last seven months I made 503 Git commits.

Last Week’s Spike

At the end of June I started talking with a user who wanted access to a JSON API to convert a lot of documents. They also wanted results to be normalised like this:

    "raw": [
            "date": "03/08/20",
            "description": "Monthly Service Fee",
            "credit": null,
            "debit": "5.00",
            "balance": "777.07",
            "amount": null,
            "charge": null
    "normalised": [
            "date": "03/08/20",
            "description": "Monthly Service Fee",
            "amount": "-5.00"

It took a lot of work to support all the different bank statement types for this customer, but eventually I got it stable and accurate enough for them. Last week this user signed up and started a large recurring payment.

Improving Conversion Results

A large portion of those 503 commits was spent on improving conversion results for various banks. If you come along and upload a Regions Bank statement and we failed to process it properly, you’ll close the app and never return. If we do it correctly, you’re very likely to buy a subscription.

Doing this mole whacking sort of work is pretty tedious, but it has an impact on sales.

Next: Localisation

At the moment the majority of sales revenue comes from English speaking countries. I suspect this PDF bank statement problem is a worldwide problem, so why don’t we have sales in Mexico, China, Japan or Germany? I have a feeling it’s because our website is in English and people in those countries will be searching in Spanish, Chinese, Japanese and German.

To test this ‘suspicion’ I’m going to localise the website and blog into Spanish, Chinese, Japanese and German.

Next: Enterprise Sales

I suspect there are large organisations out there with employees manually going reading through PDFs for various reasons. Here’s a few I can think of:

  • Tax departments
  • Financial institutions
  • Insurance companies
  • Mortgage brokers
  • “Know Your Customer” providers
  • Accounting software companies
  • Bank Record Fetchers like Plaid

I’ve never done ‘enterprise sales’ before, but I’m sure I can figure it out.

Next: Xero Integration

A lot of my users use Xero. They use Bank Statement Converter to get the transactions out of a PDF and then upload the CSV into Xero. Then they do whatever else they need to do with Xero.

I think accountants would prefer to use a Xero plugin that allowed them to upload PDF Bank Statements and have the underlying transaction data in their Xero account.

Next: Talk to Users

YCombinator love telling entrepreneurs to talk to their users. I do all the customer support, but I don’t actively contact users. I will start doing that.

Next: Feedback Form

A lot of users try the app anonymously, and then drop off. I believe this is because something went wrong with the resulting CSV. I plan to include a small feedback widget in the results page so users can tell us if anything went wrong with the conversion.

See ya

See you in a few weeks when I cross the $10,000 MRR mark

Join The Mailing List