October 24 2021
How I Managed Cash Flow at a Bootstrapped Startup
Retrospect, Inc. was a bootstrapped startup of sorts. We were spun out in 2011, responsible for a product that was 22 years old, with around thirty people at the company. We had an accounting firm, Accretive (now Countsy), with a fantastic and helpful advisor there, and they maintained our financials in NetSuite. From 2016 until our acquisition in 2019, I was Chief Operating Officer (COO), so I was responsible for the financials and the cash flow.
The cash flow of a small company is composed of two pieces: Accounts Payable (AP) and Accounts Receivable (AR). AP is the list of vendors to whom we own money with our payment deadline. AR is the list of customers that owe us money with their payment deadline. Sales closes opportunities (bookings) and hands off booked orders to Operations; Operations fulfills those orders and bills customers (billings), transitioning those orders into AR and over to Finance; and finally, Finance follows up on AR and collects money from customers for those orders (collections). This back-office process is how product turns into cash, and, for us, it involved many manual steps.
The core component of a bootstrapped startup is cash in the bank. We needed to know our current financial position, predict the company’s cash flow, and minimize past-due AR. Let’s walk through how we did that with a simple spreadsheet.
Predicting Cash Flow
I’m not referring to complicated machine learning models to predict cash flow. We needed basic spreadsheet-driven estimates for our expenses and our cash. We did not have a subscription service at that point, so we did not deal with annual recurring revenue (ARR) or subscription churn rate. Our focus was simply understanding our current business, selling software licenses online and in the channel as well as support contracts and collecting payment from customers.
AP was consistent for us every month. Countsy did a great job maintaining our NetSuite instance, so that I could drill into any department and see their expenses for the three years or the last quarter. For us, the largest expense was payroll. Beyond that, we had sales rep contractors, our European support team, our accounting team, and a long list of credit card charges. Every quarter, we had to account for quarterly bonuses as well as quarterly bills, such as Salesforce. When we grew our Sales staff, I knew how much those additional people were going to cost on a monthly basis. There were occasional out-of-the-blue bills, such as the yearly online store fee, but in general, our burn rate was easy to predict.
AR took more effort. Most of our customers had NET30 payment schedules, meaning that they had 30 days to pay us after we sent the invoice. However, when I started as COO, we had an AR forecast of one week, meaning we knew which customers were going to pay us what amount for the next seven days. Think about those two data points. Our customers had 30 days to pay us, and we only predicted payments a week out. We had better data than a one-week horizon.
To predict cash as far out as possible, I built an “AR Forecast” spreadsheet in Google Sheets. Using a report in NetSuite, I copied every customer’s open invoices with amount, purchase date, and payment terms into the spreadsheet, and I used formulas to note when the invoice was due. For each customer, I learned what their past payment schedule was and manually predicted around when they would pay us. The spreadsheet allowed me to extend our AR forecast from 1 week to 6 weeks as well as accurately track our total AR.
As a bootstrapped startup, tracking AR and predicting payments were critical to understanding our runway and spotting any hurdles.
Minimizing Past-Due AR
In addition to only having a 1-week cash prediction, we had a huge past-due AR balance. According to NetSuite, many customers had unpaid invoices that were over 90 days old. Some open invoices were over a year old.
All customers had their own processes for paying vendors. Large distributors’ Finance departments ran like clockwork, but they were still idiosyncratic. For instance, one consistently sent checks every week for all of the open invoices that had past NET30 that week, but the key phrase was “sent checks”. The checks were marked as sent in their system, but they always managed to arrive two weeks later. Adding up the times, the checks arrived six weeks after we sold the products: NET45, not NET30. Still, large customers were consistent.
Smaller customers varied in terms of how prompt they were at paying. We had a number of small distributors that were great at paying on time. Others needed reminders. One change we made was ensuring there was an employee who was responsible for following up on past-due AR.
A past-due invoice did not necessarily mean that we hadn’t been paid. All customers need to submit their payment with remittance information, which lists what invoices the payment covers. Sometimes, remittance information is incorrect, leading to accounting not closing the correct open invoice. Other times, accounting has the correct information but has not yet updated the open invoice.
Our huge past-due AR balance was a mix of all three. Accounting only processed invoices every other week, so there were frequently unprocessed invoices. A subset of remittance information was incorrect, so while the customer thought the invoice was paid, our system did not. Working with customers resolved a long list of these that had built up over a couple years. Finally, a couple customers were consistently late at paying. Our diligent back-office person worked with them to figure out a payment schedule that we both agreed on.
By minimizing our past-due AR, I knew what our real AR was, so I could predict payments more accurately.
Profit: EBITDA vs Cash
We didn’t need to worry about EBITDA–shorthand for earnings before interest, taxes, depreciation, and amortization. It’s a common accounting measure for a company’s financial health. As a bootstrapped startup, cash was critical, and EBITDA was only useful for our corporate taxes. We followed cash basis accounting (as opposed to accrual basis accounting), and we focused on profit in terms of cash to decide questions like hiring, bonuses, or layoffs.
Understanding Data and Processes
Understanding a small company’s cash flow isn’t hard, but it does require patience and tools. NetSuite is a fantastic back office tool, but it’s not necessarily ground truth. We couldn’t simply click on “AR Report” and accept the sum at the bottom. The data in the report wasn’t accurate in multiple ways. Digging into the actual data, customer by customer and vendor by vendor, helped us see what our actual AR was, allowed us to predict our cash up to six weeks using a simple spreadsheet, and enabled us to manage the cash flow for our bootstrapped startup.
