Cash Management In Startups
When I was in my mid-20s and had just gotten a job in venture capital, I read a piece on Alan Shugart, the larger than life founder of Seagate, one of the most successful disk drive companies. Alan was quoted as saying that “cash is more important than your mother.” That got my attention because mothers are really important.
Over the years, I have learned what Alan meant. Cash is everything in a startup. It is the fuel that keeps the car running. And startups fail largely because they run out of cash.
I was working with one of our portfolio companies yesterday on a cash forecasting model, a practice that I strongly recommend and appreciate greatly.
Forecasting cash is more art than science, particularly in a growing company where there are all sorts of unpredictable things (revenue, infrastructure costs, hiring pace, receivables, etc).
But like all practices, it is about the practice. You have to engage in cash forecasting, you have to engage in it regularly, you have to adapt to changing conditions on the ground, and you have to internalize the puts and takes and their impact on operations.
When a company gets mature in their business operations, has repeatable revenues, and has a strong balance sheet, you can run the business based on an annual plan or a semi-annual plan.
But early on in a company’s life, you are going to have to operate on an ever changing plan. If the revenues are coming in more slowly, you hire more slowly. If you want to wait another six months to raise more capital, you have to buy that time with changes to the operating model.
That is the back and forth between cash management/forecasting and operating. They go hand in hand.
It all starts with a cash forecasting model. It looks like a forward-looking profit and loss statement. But you do it on a cash basis. And you include balance sheet items like security deposits, equipment purchases, etc in it. Think of it as forecasting your checking account over the next year.
Once you have that, you need to engage in updating the model regularly and it is ideal to do that as a team, or at least with parts of the team, in the room. That makes it abundantly clear to everyone how operational decisions impact cash and runway.
I like a weekly cadence to this process and I like it to happen with the key senior leaders in the room. That may feel like wasting people’s time. But cash is more important than your mother in a startup, so managing it is never a waste of time.
Good one.Running marketing and sales cross tiny startups and public companies, where my groups out of pocket was always the companies largest, I did budgets every month for decades.Sitting down with the CFO or early on a board advisor was a perpetually clarifying and focusing and grounding process.You are simply right on.
Learning how to manage cash—or perhaps more accurately, “paying for it” for the times I didn’t manage it well—has been one of the most impactful lessons from running Orbital.
If you see your way to writing about a specific topic, maybe you could give us some thoughts about managing our actual mothers tomorrow during Thanksgiving. That’s something I would love to read. 🙂
.Mothers are easy. Talk to them. Answer their questions.At every suggestion they make: “Yes, ma’am.”Then tell them in front of everyone how blessed you were to be their son.I wish my mother was alive so I could do that.JLMwww.themusingsofthebigredca…
Thanks Fred — very important post. I have been forecasting cash flows weekly for a year as CFO for a PE portfolio company. My process is decent; however, if you have any files you could share for a best practice format that would be much appreciated. I am convinced I can improve our process with some help.
Very good post Fred and so true, thank you. Might be useful to share an example if you happen to have one available.
You’re spot on Fred.One of the ways I do this in a startup is to avoid the “cost center” budgeting that we often inherit from the big companies we work at. Instead I get the team to build one, big, flat, cash model. It starts with “cash-out” at the top: labor, marketing, travel, rent, SaaS fees, fixed COGS like AWS, etc. That’s followed by a forecast model for “cash-in”: recurring and one-time revenue, converted to cash payments (not GAAP revenue), less commissions and any variable cash-COGS. When it gets fancy, the team drives the revenue from CRM, with formulas for payment timing.This is easier than building a rolling pro forma income statement. And you can still delegate ownership of spending. My suggestion is to avoid building the big “department” budget model, and instead drive a weekly cash model that you look at before every leadership team meeting.
Love this piece, Fred. As someone who works B2C, we work with individuals to build their own “trend reports.” It’s something most folks do not do with their own cash flow (but should). Using their own personal trend report helps them often dodge trouble.There’s a deeper message in what you’re saying – in that forestalling, or deferring purchases or investments usually works out better, than rushing ahead.Again, many thanks. Tom
I use my cash forecast as my business planning model in “live” mode i.e. I use and update it weekly – I also combine PNL and Cash based accounting to keep the numbers real and match with the bank statement every 2 weeks. Its tough to match up numbers weekly because of the way debits occur for payroll etc – But it becomes more realistic if you budgeted $50k in burn and actually spend $70k in burn because you had some incidentals and you see it over 3 months and realize thats your actual burn.. Not $50k.Accountants are useless here because classical accounting doesn’t help at all – except *maybe* with AR forecasting to show who you need to push for receivables.I keep my model on Dropbox and share it with investors so they can “Dip” into it multiple times whenever they see fit.I also create projections and snap shot them by versioning them for board meetings 2X a year so I can actually go back to those projections and see what the assumptions were and where they changed (they always change)After 6+ years of startup-ing I have a model that shows both ATL and BTL numbers and is operable for Operators.The Sales numbers and churn are linked to a Sales Model that models out churn and growth and I plug in actual numbers every month to update that model as well.
.Very strong approach. Well played.This is the voice of experience — 6 years of constant improvement — that is driven by having actually done something.The fact that all of your work results in an actual work product should not be overlooked.This is why the risk factor of investing in serial entrepreneurs is so clearly diminished.Well played. Well done. Keep it up.It would be useful to train someone within the organization to do some of this to lighten your burden and carve out time to focus on the actual business.JLMwww.themusingsofthebigredca…
Thank you. Learning every day. Still Seed stage but the model is open enough for people to grasp within a few days. Takes discipline is all.
wowl congratulations Pranay. What a nice job.Does it require auto learning?
Really good insights here. Would you be willing to share the model/doc? Even a watered down version would be super helpful. Thanks.
It’s cash or crash.More important than your mother? From a man who tried to get his dog elected to Congress. That’s not even eccentric.
Very good advice, and founders who have lost a startup due to poor cash management, or who have experienced a serious cash crunch but survived, know this lesson well. I might even say that there are two types of founders, those who have learned from a cash crunch, and those who will.
I forecast cash flow a little bit differently. My focus is on forecasting all the non-cash items on the balance sheet, which obviously gives me the cash balance in order to have a balance sheet that balances.The income statement forecast automatically feeds into the balance sheet forecast. There are linkages such as net income and equity, revenue and accounts receivable, depreciation expense and fixed assets, etc.The advantage of such a model is that it is easier than trying to develop a cash-basis income statement and then trying to adjust for changes in the balance sheet.
I have found adopting a rolling 13-week cash flow forecast to be helpful in both distressed and non-distressed situations. Reviewing it regularly with the management team is so much more enlightening than doing a monthly review of the standard P&L and BS.
Important topic to any business. I think a simple variant of a Statement of Cash Flows can be helpful, or a Sources/Uses worksheet. Revenue is not cash, neither are expenses. Always looking for clever new models to employ.I expect a good CEO will always know their current cash position, general burn rate and any upcoming outlays. For manufacturing companies you have working capital considerations which adds a layer of complexity.Funny thing, I think cash constraints can force creative problem solving, focused prioritization which can result in some amazing things.One other thought. As important as cash might be, of equal importance is the ability to make measurable progress over time toward specific goals for the business – growth, profitability, learnings. Measure the effectiveness of your investments i.e. are you getting good value for the money you are spending, advancing to the next level. For a start-up business the plan never, never works. But are you ok with the progress, what have you learned, what adjustments need to be made to improve, is your current plan still viable? This is not a weekly exercise, but the tactics of cash management support the larger strategic goals.
If want to do some cash planning under wild uncertainty, then (i) get data from USV on how much uncertainty there was historically in early stage startups, (ii) convert that data to some rough cut statistics (NO Albert, do NOT assume a Gaussian distribution; instead maybe just use the empirical distributions from the actual DATA), and (iii) set up a spreadsheet with, in simple terms, one row for each variable and one column for each month. Then use the spreadsheet feature of Monte Carlo simulation. Then for some decision options, get the probability of going bust if spend now instead of save. For more you have, surprisingly, essentially formulated, precisely enough to be attacked with real likely relevant data results by the applied math of stochastic optimal control (disclosure: I got a Ph.D. in that field).Here is a simpler approach I am using and intend to use at least during early growth: (1) If I spend prudently, my checkbook has funds enough to last and meet all the routine expenses of my startup through great revenue (if that happens). So, no chance of going broke now! (2) As grow, never but NEVER count chickens before they hatch. Or, don’t spend money until it’s in the bank with taxes, etc. accounted for. (3) For taking on continuing expense obligations, e.g., leases, employees, follow the Bill Gates Rule (based on a rumor): Have cash enough in the bank to meet those obligations for at least 1 full year with no, none, zip, zilch, zero, nichts. nil, nada revenue at ALL. (4) While my startup is way up over the tops of the charts, both in practice and in principle, 100% totally squeaky clean, uplifting, safe for work, families, and children, culturally uplifting, good for civilization, STILL stay the heck out of sight of any and all poisonous snakes, rabid bats, black death rats, incurable flesh eating bacteria, highly toxic hazmat Super Fund sites, high level radioactive waste, or lawyers — all such things DO exist, and they are NOT my fault. Look into making continual generous contributions to, along with smelly confidential gotcha files on, relevant politicians, judges, and police. Once the business is at all significant, for everything of value, HIDE it, using LLC, C-corp, subsidiaries, shell companies, off-shore whatever, etc. Be prepared instantly just to make a public announcement on the Web, then shutdown, move to a different country, open again under a new name, make another announcement, and continue.Part of this is to have a startup that CAN do such things: For that, want a problem to solve, say, that 3+ billion people very much want solved everyday. Have by a wide margin the uniquely best world class solution. Have some high barriers to entry. Have a business such that the operating expenses and capital expenses are down in the rounding error of the revenue, a business that can be started and brought to, say, $20 million a year in revenue by just one sole, solo founder, a business that continues to make money evenings, weekends, while the founder sleeps, goes to symphony concerts and seminars on math and physics. The keys to doing that? Exploit the astounding price/performance of current computing and the Internet. Get some new data. Process the new data with some powerful, valuable, difficult to duplicate or equal, rock solid (NO data science, machine learning, or artificial intelligence in sight) original applied math to solve the users’ problem. Then the good results go to the happy users. Nothing else has even as much as a weak little hollow hint of a tiny clue how to compete.
.Statement of Cash FlowsSometimes we make things way harder than they need to be. I ran companies — public and private — for 33 years and it took me 5 years to learn how to forecast effectively. [Outliers, Malcolm Gladwell type experience]As it relates to running any business, the first thing is to ensure the tools are in place and they are easy to find, configure, and use. They’re all in Quicken.Here is what every company has to have:Chart of AccountsGeneral LedgerCash Receipts JournalCash Disbursements JournalCheckbookCredit CardsIncome StatementBalance SheetStatement of Cash FlowsI listed the Statement of Cash Flows — which is the most often ignored financial statement — last, because it is the one that begins to solve the problem you lay out. Understand that the SCF takes info from the Income Statement and Balance Sheet and melds it into the cash flow of the company. If you focus on the Income Statement, then you only see a part of the cash flow.Very few people in business understand the SCF. It is the totality of the movement of cash — cash to/from operations, cash to/from investing activities (capital transactions), and cash to/from financing activities. [When you run a GAAP reporting company, you can even add in “non-cash” items to get a GAAP picture.]As generated by an accounting system, it is a “historical” document. As used in your example, it is a forecast and provides a road map for the future.To build the prospective model, it is important to attack each of the constituent elements (operations, investments, financing) in the same prospective manner. In essence, you are forecasting each of these elements. [If you raise VC funding or a subsequent round, you are just updating the “financing” — proceeds from financing — section of the Statement of Cash Flows.]Understanding these components makes the job much easier as you can compartmentalize your forecasts, delegate some of them, and see where your most accurate and least accurate elements are coming from.A couple of useful tricks:1. Graph everything you can. The trend is almost inviolable. It is real. You will not get results that are wildly off trend. It is trite to say, “Make the trend your friend,” but it is true. When the company begins to grow or decline, it will show up on the trend line immediately.2. Make three forecasts — high, medium, low.3. Over time see where reality and your forecast link. I always used to think the “medium” forecast was also the “most likely,” but my actual experience was that it was the “high” forecast that turned into reality. That was for me. Yours will be yours.4. Publish the methodology widely. Make sure everyone knows how it was made and the constituent parts. Why? Because you will find out there is one single area of the forecast that you are getting wrong — for me, it was always the capital expenditures. I would buy stuff without thinking about the cash flow impact.5. Delegate what you can. The graphs are an area that a CEO should not be doing.6. Hire CFOs who can do this stuff and who are organized. As a CEO, your objective is to delegate this at the requisite level of craftsmanship — completely.I had a long serving CFO who was like a wizard in his ability to forecast. We did monthly forecasts because we had no cash issues, but his forecasts were so accurate that I would often splash Holy Water in his office to ensure he wasn’t in league with the Devil.7. Keep all of your forecasts and look backwards comparing the actual future with the forecasted future. When I was growing like mad, I could see how unrealistically conservative I was. It was a damn good lesson.8. If you are a small company and cannot afford a CFO type financial hand, find a VCFO — virtual chief financial officer — company and find a guy with 30+ years of experience to look over your shoulder. In a very few hours, such a seasoned hand can help you. These kind of outfits are springing up all over the place.The more you do this, the better you will get until you can “feel” the throb of the cash flow heart, but you can’t do it if you ignore all the necessities of the accounting system.Last point — make damn sure you know what cash is in the pipeline. This is why a cash receipts journal, a cash disbursements journal, the checkbook, the credit cards, and the bank statements have to reconciled on a regular basis.Today, you can look at bank accounts on line in real time, so you can do this on a weekly basis. Do not overlook billed, but unpaid credit card commitments. Credit cards are always a problem.I have literally seen startups go from a shoebox of invoices, a checkbook that has never been balanced, personal credit cards, actual cash into precise Prussian order in a couple of months. Thank you, Quicken.Like a lot of things — y’all did not invent sex or business. The tools to manage cash flow, to predict cash flow have been baked into the cake for a long time — Statement of Cash Flows.JLMwww.themusingsofthebigredca…
“That got my attention because mothers are really important.” Made my day hahaha. I guess when CFs are not significantly negative, breakeven, or slightly positive, founders have much better sleep at night.
There’s also this quote from Harvard’s Bill Sahlman: “Startups are planning to run out of cash.”After losing a bunch a money, here’s what we teach founders to do:- finalize your cash flow forecast in the first month post-closing, if you don’t have one. Get it audited by professionals.- hire a finance person to oversee the process (as soon as post-Series A). Not an accountant. The person needs to be forward-looking, accountants generally aren’t. Investment in the finance function is one you will never regret.- update the statement monthly and do current month actuals + next 2 months. You know what you billed so you can assess cash-ins.- graph a curve that shows end-of-period cash balance excluding new business. That’s your potential lowest cash point.Money in the bank is the resource that vanishes the fastest post-fundraising.
Cash flow is one of my favorite topics and as my 4.5 year old likes to say when we visit the bank – money-in, money-out…it’s not the other way around.I counsel others to avoid getting too comfortable with a successful business cycle or frothy bank balance – never be lax about cash flows. Always stay on top of it and master your understanding of all the operating levers to cash management and efficiency. Eventually, a time will come when you feel the squeeze, and your understanding of operational effects in how money-in and money-out works, will greatly improve your success.
As a CFO for 20 years, often with early-stage companies, I could not agree more. Regular communication is especially important. Management and investors should never be surprised by the company’s cash position and remaining runway.Lots of good practices and suggestions mentioned in these comments. One thing I’d add is that I don’t use stock accounting reports (e.g., statement of cash flows) for projected cash flows. Whether it’s an annual, monthly, or rolling 13-week cash flow, I lay it out in a logical fashion, with commonly-used terms, that’s easy for non-accountants to understand.
Al (seed), then you via Rex @ SoftbankVC, gave me my big break. I’m forever grateful for everything I learned from you guys back then
Thanks Fred. We had this problem – and decided to build a solution for it! It’s relatively easy to get a statement of cashflows if you keep your accounting software up to date. Especially if you automate the bookkeeping with a solution like ReceiptBank or HubDoc. We found that it was managing all the expected payment dates on invoices and bills, plus setting future “budgets” for cash coming in and out. Don’t know if you’ve seen Float? But would love to get your thoughts on it.