Projections, Budgeting and Forecasting

MBA Mondays is starting a new topic this week. It's a big one and I think we'll end up doing at least four and maybe even five posts on this topic in the coming weeks.

I said the following in one of my first MBA Mondays posts:

companies are worth the "present value" of "future cash flows"

The point being that the past doesn't matter too much when it comes to valuing companies. It's all about what is going to happen in the future. And that requires projecting the future.

There is another big reason why projections matter. They are used for goal and expectation setting. Generally speaking goal setting is used to manage the team and expectation setting is used to manage the board, investors, and other important stakeholders.

And finally, projections matter because they tell you what your financing needs are. It is critical to know when you will need additional financing so you can start planning and executing the process well in advance of running out of cash (I like 6 months).

There are three important kinds of projections. I'll outline each of them.

1) Projections – These are a set of numbers, both financial and operational, that you make about your business for various purposes, including raising capital. They are aspirational and are often done with a "what could be" perspective.

2) Budgets – These are a set of numbers, both financial and operational, that the management team prepares each year, usually in the fall, that outline what the company plans to achieve in the coming year. They are presented and approved by the board and the management team's compensation is often driven by them.

3) Forecasts – These are iterations of the budget that are done intra-year by the management team to indicate what is likely to occur. They reflect the fact that the actual performance is going to vary from budget (in both positive and negative ways) and it is important to know where the numbers will actually end up.

Over the coming weeks, I will go through the processes companies use to project, budget, and forecast. Because I do not do this work myself, I've enlisted one of our portfolio companies to help me with these posts. 

I've been working with Return Path for ten years now. Matt Blumberg, CEO, and Jack Sinclair, CFO and sometimes COO, have done over ten sets of projections, budgets, and forecasts for me and other investors, board members, and team members. In the process they have evolved from a raw startup to a well oiled machine. With their help, I will talk about the how three "model companies" go about projecting, budgeting, and forecasting. These companies will be 10 person, 75 person, and 150 person. These are the typical sizes of companies that I work with and are probably also the sizes of companies that most of the readers of this blog are dealing with.

I'll end this post with a picture that Matt sent me last week. This is ten years worth of board books that include Return Path's projections, budgets, and forecasts. The goal of MBA Mondays in the coming weeks will be to get all of you to a place where you can create something similar.

Rp budgets

Reblog this post [with Zemanta]

#MBA Mondays

Comments (Archived):

  1. RichardF

    I think many entrepreneurs just starting out are going to find this really useful.Love the three models concept and the fact that you have enlisted a portfolio company to help you.

    1. fredwilson

      yeah, i was worried that i don’t have enough operational experience with theprocess of putting together these thingsi see the end product all the time, but the “sausage making” is what i wantto get into

      1. RichardF

        As a former bratwurst maker I’ll be watching with interest!

        1. fredwilson

          are you serious about making bratwurst?

          1. RichardF

            No – I meant the number crunching – sorry if I got your hopes up!

          2. fredwilson

            you did, but i was also pretty sure you meant numbersi look forward to your comments since you’ve been through this

          3. Carl Rahn Griffith

            Richard, old chap – you know the rules – number crunchers are not supposed to make jokes!;-)

          4. RichardF

            that’s why I’m ex-number cruncher!Also it’s one of the answers to your question about why so many start up’s avoid it – it’s not sexy

          5. Carl Rahn Griffith

            Lol, fair point.Indeed. Shame so many can’t see that running out of money/going bust is a hell of a lot less sexy!

          6. Tereza

            Uggh. Running out of money…..TOTALLY unsexy.

          7. Mark Essel

            Haha, I’ll second that Richard. Although we need to be intamitely familiar with capital into and out of the business. I’ve got a long runway with part time income.I sell 30 hours a week to my day job (I think of my time as a company resource), and my cofounder has the same setup (although he’s a better hacker by far than I). Development takes longer than I like. I think slower at night and am learning a bunch, but days off are perfect. Once we converge on a functional design (many direction changes between last October and now) I’ll be much more comfortable “burning the barn”, certainly if we can land a working business model (income).

          8. Tereza

            I say, start with big numbers and a very hungry customer, and work backward from there.People are delighted when you give ’em what they asked for.

          9. markslater

            i am an ex cruncher as well! i actually thoroughly enjoy modelling even today.In my opinion – the forst 24-36 months are where the most focus should be placed – anything beyond that is guess work.Its also important to use sensitivities in the early stages – if you are pre-revenue then likely your business drivers will have hidden sensitivities to your model that can prove an Achilles heel.

          10. RichardF

            I agree Mark once you start getting beyond 24 months you are sticking your finger in the air.

          11. Tereza

            Yeah I’m with you. I’m always the words girl but loooove seeing the sensitivities reveal themselves because they so thoroughly define not only the words and the positioning, but the operational actions.And THAT is sexy!

          12. Tereza

            um….if the numbers are big, they’re sexy.:)

          13. RichardF


          14. JLM

            If you like to shoot deer, as I do, you have to know how to make deer jerky, summer sausage and hot spicey links.A lot of venison (except for the tenderloins which you marinate, grill and finish in the oven), a bit of beef and a fair amount of pork — lots of spices — ground finely together and stuffed into a skin, soak in ice cold water and then smoked until just barely wrinkled. Age them for a month, store some in the frig, freeze some and then eat w/ sharp cheddar and peppered water crackers.Until next deer season.Bambi, I love you!

          15. Tereza

            We have a family of six deer that frequent our property and eat all my plantings.JLM if you want to have at ’em, be my guest.But the rule is, you have to take a doe before you take a buck.

          16. RichardF

            I love venison in any form (those sausages sound delicious!), in fact I love game in any form….wait are we on the right Wilson blog? we should probably be discussing it here – .Only been stalking in Scotland once and Bambi didn’t turn up, however when the season starts again I’m hoping Donald (Duck) won’t have headed South.

          17. JLM

            I love Scotland. Great fishing, hunting, golf, riding and countryside. I just can’t understand the language — what language do they speak in Scotland? Just kidding. I know it’s Anglish.A million years ago in the mid 1970s before most persons on this blog were a gleam in their Daddy’s eye, I used to slip down from the DMZ in Korea in my jeep dressed in fatigues to the Naija Hotel in Seoul. Strictly forbidden! There was a Ghurka battalion which had Scottish NCOs. My last name is the same as the seas off Scotland and they were convinced that I was Scottish. It used to take 10-12 rounds of drinks to work through the problem. The ride back was always an adventure but I was young and dumb and filled with …

  2. Guest

    “The goal of MBA Mondays in the coming weeks will be to get all of you to a place where you can create something similar.”Something similar hopefully being ten years worth of board books – that would definitely be something!

    1. fredwilson

      yes, any entrepreneur who can point to ten years of board books has achievedsomething for sureand that is part of the goal of MBA Mondays – to make entrepreneurs moresuccessful

      1. Keith B. Nowak

        Thank you! The advice and help has been invaluable.

  3. Dan Ramsden

    The nice thing about this way of presenting the subject is that it splits the financial forecast into (a) long term (projections), (b) short term (budgets), and (c) real time (forecasts/adjustments). For digital media, I continue to find (a) an impossible exercise, and it will be really interesting to see if discussions in coming weeks begin to converge upon (b) and (c) mostly.With that long stack of books from 10 years worth of financial planning, it should be possible to determine the true viability of (a) based on actual results. Maybe it’s wrong to dismiss long-term projections… I would be curious. (I don’t mean whether long term targets were realized, but rather if the business that the company thought they would be in ten years ago, has turned out to be their business.)

    1. fredwilson

      i really like that way of putting it; long term, short term, and real-timei’m going to start thinking about it that waythanks!

      1. Dan Ramsden

        Thanks, Fred.

      2. ShanaC

        Much better titles too. They should rename them that.

    2. Mark Essel

      My instincts are that long term for startups is more creative than functional but your emphasis on model has me curious to see how predictions may have shifted over time

      1. Tereza

        The long term fuzzier stuff is to make sure you’re not closing off any decisions that may be important trends. Don’t want to get blindsided when it’s too late.You can only decide when the trend/opportunity is actionable if it’s on your list of possibilities.And then you can scope it small, large, core or tertiary. Or non-action.Again just want to make sure it’s a choice, not something you’re forced into.

    3. awaldstein

      Dan-great approach.This maps to my much less sophisticated Black (must have expenses and short term revenue); Grey (like to have with more expense, more risk against revenue) and White (Pie-in-the-sky…never use but good to dream).Somewhere in there, I find an operating method that works.

      1. fredwilson

        Shades of grey!

    4. zackmansfield

      Dan, I agree with others in your classification of Long term, short term and real time. I’ve worked as a venture lender and thus see dozens of forecasts/budgets a year. I often find that companies run into trouble when they mis-match and wrongly emphasize one (long, short, real time) vs. the other – i.e. the long term plan calls for a big ramp up in headcount to meet future growth…which doesn’t materialize and is not addressed in real-time quickly enough. Next thing you know you’re 2 months from a cash out date.I think the opposite can be true as well – management is so caught up in the weeds of the day to day – winning the specific deal or finishing the particular feature – that there is not thought as to how/if this plays into the longer term plans.

      1. fredwilson

        Venture lending is a great place to see the disconnect between plans and reality

    5. paramendra

      Take Tumblr that has not made any money in its few years of existence. It will still have (a), (b), (c). I think.

    6. PhilipSugar

      I like this as well Dan.I think that the viability of (a) long term (projections) is zero if you try and correlate to short term, and real-time.However….That doesn’t mean you shouldn’t come up with projections (no more than three years) because “if you don’t know where you’re going any road will get you there”By the same token projections are meant to be changed and not used as a hammer when they end up being wrong.

  4. Carl Rahn Griffith

    The incompetence, complacency and naivete of so many start-ups (and even more mature companies) with regards to this topic borders on the breathtaking.Seems to be particularly prevalent to our industry.Thoughts as to why?

    1. David Semeria

      The people who are itching to do a startup are unlikely to have the patience to learn this stuff. Especially since the subject can be very tedious if you’re not applying it to a real situation.

      1. Tereza

        May depend on the type of business.On a general basis, I agree with you.But I’ve met many awesome entrepreneurs who’ve grokked their businesses through the bottom-up model assumptions.Some way or another the entrepreneur needs to understand, on a detailed basis, what makes their business profitable, and what tradeoffs they need to make together there.If it’s very early stage and it’s something that’s never been done before, then it doesn’t feel like a ‘real situation’ and feels tedious.If it’s a business where execution makes a difference — which is most businesses — then sooner or later they won’t be able to help but get dragged in.

      2. Fernando Gutierrez

        I don’t think the problem is not having the patience to learn, but not having the patience to do. Most concepts related to forecasting are pure common sense. But the benefits of forecasting are not as immediate as many day to day tasks, so many of us procrastinate on this to attend more urgent (apparently) things.

    2. JLM

      What a great riddle.I suspect that so much energy goes into the creative thinking surrounding the product that no energy is left for the basic structure of business management.I also suspect that the 1/3 great, 1/3 get your bait back, 1/3 disaster return expectations contributes to it.I wonder if computers are making folks fundamentally lazy?There a huge number of businesses which can be made into equally huge financial successes with simple modern management, energy and modern marketing. Gobs of money can be made in these enterprises.

      1. fredwilson

        i think your second paragraph nails it JLM

  5. Mark Essel

    Awesome post Fred. The MBA Monday info is great to have a basic understanding of. But I usually dread MBA “spreadsheet” Mondays because I fear falling asleep while out walking and getting runover ;)Now based on the presale I’m looking forward to next week.

    1. paramendra

      You read AVC posts while walking?

      1. Mark Essel

        Yeah, I write a post to my blog, and catch up on any reading I want to do (I check AVC every day) while walking each morning (between 6-9am)

        1. ShanaC

          Mark- I worry for your safety sometimes. Just make sure you are paying more than adequate attention to your surroundings…

          1. Mark Essel

            Roger that! I do feel much safer on back roads with sidewalks.

          2. Fernando Gutierrez

            What device do you use to read while walking? I though you would use a phone, but if you do it for three hours a day and you also write I hope you have a bigger screen than that!

          3. Mark Essel

            It’s a phone.An iPhone 3Gs that I’ve been trying to replace for a few months with an Android. Now I’m waiting for 4G to swap.

          4. Tereza

            You’re not walking around with your iPad?LOL

          5. Mark Essel

            That baby is the living room couch browser. So terrible for writing anything though (I can type faster on the phone from a year of practice)

          6. Tereza

            Well that’s a relief.My concern, aside from you getting hit by a car or clipped by a biker, is that if you were walking on the street while typing on an iPad you’d look like a total dork.It is my civic duty to not let that happen to you, Mark. 🙂

          7. Mark Essel

            :DMy soapboxHttp://

        2. Donna Brewington White

          Duly, duly impressed! You’ve turned the saying about being able to walk and chew gum at the same time on its head.

          1. Mark Essel

            Thanks Donna. It took a little while to get used to, and I wouldn’t try it on a busy street. But walking around central park was fine while writing/reading.

        3. fredwilson

          the active blogger. i’m into that too mark

    2. Tereza

      Yeah, on Mondays I’m always afraid Fred is going to sniff me out as being really sucky at finance.OOPS! Cat’s out of the bag!But seriously, I’m looking forward to this one.

      1. Donna Brewington White

        Thank you for saying this, Tereza, smart as you are. Makes me feel a tad bit more at home. I actually love finance — although sometimes understand it more intuitively than otherwise. Love grappling with these concepts.

  6. Matthew Taylor

    I really wish there were a consensus on these types of issues among “investors.”In all honesty, valuation appears to be a state of mind for each individual with few consistencies among them.

    1. fredwilson

      i think that’s true (inconsistent valuations among investors) and it is frustrating for entrepreneursbut the best way to fix that is to run a competitive process when you are financing

  7. robertavila

    For some reason people always think someone else’s forecast is better than there own. During the “reliable sources” published numbers on growth in the internet market and consultants took these numbers and ran with them, convincing companies to bet billions on them. Having done plenty of market forecasts and knowing the problem with new markets,I called a few of these sources and talked to the forecasters. They were using the standard one point regression technique based on suspect data, BUT providing a forecast where everyone was desperate to know what to expect. We just saw the same thing in the sub-prime mortgage market with home prices. A forecast that you can understand and can defend is always better than something from a third party source based only on authority and reputation.

    1. Tereza

      Great comment, Robert.I wish I could “like” this twice.I always love your spin on where the bodies are buried.

    2. fredwilson

      For sure

  8. RichardF

    Three things that are important when thinking about financial planning:1) Ensure that every member of the management team (even if there are only two of you) is engaged in the process. Every one of them needs to understand why the process is important and their input and buy-in to the process is crucial. It really helps to ensure that everyone is aligned with achieving the company goals with the box of cash you been given.2)When you are in start up phase with little or no revenue the most important figure to focus on is cash and the length of your runway. I think pictures get the message over most succinctly. So graphing cash burn over time is a real attention grabber.3) Fred has already spoken about it – bottom up, in my opinion, is the way to make your plans. Identify your key cost drivers and revenue producing metrics and move from there. Headcount growth is something that many ‘early stage’ VC backed start up’s fail to get a handle on. Just after raising a round there can be a lot of pressure from existing employees to recruit reports.In Europe particularly, headcount cost is not something that can be quickly removed once you have incurred it. Notice periods and contractual termination payments mean that will not see an immediate reduction in cost if you need to cut headcount (which is usually one of the largest costs in a company). So you need to have an accurate idea of what your recruiting plans are.

    1. markslater

      i have heard a 100 start up say – “if we could grab 2% of the market” and i’ve never found this useful. Bottom up all the way

      1. Greg Gentschev

        I agree, in general. There are maybe two cases where top-down can be useful. One, when the top-down is smaller than the bottom-up. Then it forces you to reassess your revenue per customer assumptions, which could be too optimistic. Two, when you realize another part of the market you sliced away in your top-down sizing is significantly bigger than the one you’re left with. That should prompt some thinking about whether you’re really targeting the right segment or whether you should go for a bigger slice of the pie.

    2. fredwilson

      Great advice. I second all of those points

    1. fredwilson

      Excellent. Thanks for sharing the links

  9. iTbay

    for forecasts, it best to do a best case, budget case, and worst case forecast.

    1. paramendra

      Hello there, stranger.

      1. iTbay

        Hey! We meet again in the cyber world. 🙂

    2. fredwilson

      I see forecasts as much more about narrowing down onto what will happenI like doing the various cases in the projections

      1. iTbay

        your right! I go the two mixed up. In Canada, we have an accounting standard for this in “SECTION 4250 future oriented financial information” of the Generally Accepted Accounting Principles. I am not to sure whether there are IFRS related to this stuff. Canada is adopting IFRS in 2011.

  10. Steve Koss

    Fred…please include scenario planning in the mix…cheers, Steve

    1. Fernando Gutierrez

      I did some scenario planning during my years at a big telco and also in a consulting sting I did before that. I found it specially useful to forecast things with very limited information (like new businesses).Thanks to Excel and its scenario planning tools I mixed it with some statistical analysis and was able to obtain some interesting information and sensitivities. I started defining the business model. Then I would choose the key independent (this is very important) variables and define several scenarios for each of them (same number for each variable and equally separated). Then Excel would run all the possible combinations and I would be able to analyze them.This kind of approach has to be taken carefully because you can lose perspective of you don’t understand all the implications, but is great for sensitivities. And people loved when we explained that we had run 25k scenarios and that the business would be ok as long as we kept x variable above x value.

      1. Tereza

        That’s awesome.You and Robert Avila should talk. This is right up his alley.

    2. fredwilson

      I will do that. I like to do the scenarios in the projection phase

  11. Venkat

    Fred: Can you talk a little bit about how valuation practices and discussions differ between broader private equity (PE) sector investing vs. VC investing? The basic math is obviously same/similar, but the uncertainties would be very different, and I am curious about whether this shows up in leverage models etc.?

    1. Greg Gentschev

      Traditionally, PE buyouts focused on low risk businesses with stable cash flows (things like consumer packaged goods and manufacturing) and for barriers to competition. Since the business risk was typically low, they could use a lot of debt for the acquisition (say 60-70% leverage, with only a 30-40% equity investment) and increase the return for equity investors. If you buy a business for $1 and sell it for $2, that’s a 100% return. But if you only put down $0.30 in equity and financed the rest, then it’s close to a 670% ($2/$0.3) return, less interest.Recently PE firms have bought much riskier businesses, partly due to competition for good acquisitions and perhaps also due to overconfidence. A few of them have gotten burned, and there’s lots of speculation that PE will shrink similarly to how people expect VC to shrink to a more sustainable level.Anyway, the PE model is almost the opposite of VC in many ways. Low business risk/high leverage versus high business risk/no leverage.

      1. markslater

        they are also dealing with predictable projections and forecasts – they can use tools like DCFs to assess the opportunity. Using a DCF in a startup is non-sensical really as your discount rate is impossible to accurately gauge.

      2. ShanaC

        This is one of a me sort of question: As manufacturing shifts to a more lightweight business (sort of it’s easier to make things, they are lighter, it costs less, they are cheaper, hence less cash flow, those sorts of issues)- how will PE adjust to taking on service business beyond “getting burned”And how will PE develop to say a mature tech company (even though tech moves swiftly?Always wanted to know that….

        1. Greg Gentschev

          Lightweight manufacturing wouldn’t necessarily be less attractive because the value of the business just shifts to intellectual property and branding. Cash flow wouldn’t necessarily go down. Less cash might come in, but less cash would go out for investments in plant and inventory. If anything, a business that can produce more revenue with less assets and thus less capital invested would probably be attractive to any investor. Kind of like the capital efficient software startup idea that’s been going around.PE has also already gotten into the tech world. Sungard Systems was acquired for $11 billion a few years ago, and some firms like Silver Lake Partners actually focus primarily on tech (both software and hardware). A mature tech company is actually a great fit in many cases because maintenance or subscription revenue is often very predictable, as long as no disruptive innovation comes along.

          1. ShanaC

            Thank you- now the next question – that gray middle area that second marketis trying to fill- where does that fit into this….

          2. Greg Gentschev

            It’s pretty much apples and oranges. The archetypal PE investment takes an established company that’s been run inefficiently and fixes it up to get a better price down the road, both by cutting costs and trying to build on competitive advantages. Adding on a lot of debt juices the returns. But PE firms typically shy away from any significant market or technology risk, which are the staples of VC. Even when individual investment firms run both PE and VC arms, they have different managers and look at very different opportunities.Having said that, a late-stage VC investor like DST is sort of off in limbo. Investing in Facebook at this stage doesn’t really bring much market risk, nor do they fiddle with the company like many PE firms do. In practice, the distinctions aren’t always as clear as I make them out to be.BTW, PE is sometimes used as a general term that encompasses VC and hedge funds, but I think of it as purely buyout funds.

          3. ShanaC

            No I know that- Second Market is just making things screwy. If you stayprivate so long- you trade shares somewhere like Second Market or some sortof Dark Pool for these things. Some people will be involved in Managing,some will not. There still is room for innovation, but it isn’t nearly aswide as tiny company when it was first established, not further down theline if it goes IPO. So said fund that deals only in that stage is a??????? (if there is such a thing, and I expect that to develop because ofthe surity, the stage of life, and I expect that returns are going to lookdifferent/better/more complex and help get needed liquidity to really earlystage shops and help improve growth for things that need to get big.) They’re not loading on debt (yet- I mean they could try for a loan, I’m notsure if the public market will allow them to float a bond)It seems like the right time for a hybrid fund. Really involved in themanagement-really involved in late stage growth of young companies as theyget to maturity before they go to IPO- give them breathing room.That would be an interesting job.

        2. Mark Essel

          Great question Shana, was curious of how this combo was perceived.

      3. Venkat

        Thanks Greg, for a succinct and useful answer!

    2. fredwilson

      That is going to be a whole set of posts when we get thereBut to my mind VC valuations are all about properly discounting the risk so that when you do actually achieve the projections you end up with huge gainsBecause it rarely happens

  12. Pete

    There are some interesting dynamics around these processes that can cause them to break down. One dynamic is a disconnect between board or senior management expectations with reality, which pulls the numbers into fantasyland. I look forward to your thoughts on setting and managing expectations, and how management should approach these processes with board and employees alike.

    1. fredwilson

      Wow. That’s a post or two right there”We gotta show better numbers so we can get a step up on the next round””We are replacing you because you couldn’t meet the numbers in the plan”Oy

  13. Scott Carleton

    For others like me that need to look up business terminology from time to time, I highly recommend

    1. ShanaC

      Thank you. Also wikipedia gives out formulas!

      1. Scott Carleton

        With this sort of quality networking maybe an open-source MBA is possible! Especially for anyone starting a company where the title itself isn’t really required. The ones interviewing at goldman will need to forge Wharton or HBS’ seal haha

        1. ShanaC

          Dare someone here to cause one to happen….

  14. steve cheney

    I haven’t followed every comment in all of the MBA mondays, so perhaps this term has already been coined, but I thought of a great phrase for what you are doing with the MBA Mondays series:You are effectively ‘open-sourcing’ a $140K MBA – and it’s really awesome to see…

    1. paramendra

      “You are effectively ‘open-sourcing’ a $140K MBA – and it’s really awesome to see…”That is a succinct way to put it. I suspect he is also open ending it. As in, the MBA Mondays do not have to come to an end in two years or one. Most topics can be revisited over and over again.

      1. JLM

        More likely, you will never run out of things to learn over the next 20 years.

        1. fredwilson

          i learn something new every day. other than my wife and kids, i love learning more than anything.

    2. ShanaC

      I think parts should be done in an actual MBA (like in depth classwork about certain elements of your sheets) But yes. And I am having so much fun while I am doing it. And I feel like a total dork for saying so.I 100% believe in hacking education, I say after MBA monday we do book group monday of classic books that we believe entrepreneurs should read.

    3. Donna Brewington White

      Love it! Great description — open-sourcing.

    4. fredwilson

      Hacking Education one post at a time 😉

  15. Elie Seidman

    Steve Blank wrote a great post – titled “Death by Revenue Plan” about the dangers of having a revenue plan in an early stage startup. I think it serves as a good counterpoint. Curious as to your thoughts.

    1. Tereza

      You didn’t ask me, but I’ll give you mine.A lot of revenue plans are made by people who’ve never gotten revenue. By that I mean, never met with people, convinced them to do something they’ve never done before, put a dollar value to an intangible, closed deals.So there are a lot of sh*tty revenue plans out there that are quite disassociated from the market.There are also a lot that are not used correctly.Blank is correct in his description of iterative customer/product development. And if too much Board conversation is on ‘how come we don’t have revenue?’, there are a number of possible reasons, and probably a combination.Are the assumptions wrong? Are the customer conversations happening? Is the product a non-starter?These are key strategic questions which may be raised by referencing a revenue plan. And they may not be raised if you don’t have a revenue plan.So, to me, the point is, don’t misuse the revenue plan. And if it’s broken, then get out, talk to a ton of customers, re-set your requirements for the product, and write a new revenue plan that the Marketing and Sales people can actually use.

    2. Mark Essel

      I enjoyed that post, along with many of Steve’s others. He get’s down with the visuals/charts like only an engineer turned entrepreneur can :DThanks Elie, you just reminded me to check for a new post from Steve, mailed myself this one on the Lean Startups One day conference!

    3. PhilipSugar

      I don’t think Steve was saying (and I can’t speak for him) don’t have a revenue plan.He was saying don’t have a wrong revenue plan.And by wrong he was saying unrealistic.So if you haven’t figured out your value proposition with customers and you know its going to take time, don’t portray it otherwise… he points out your vc’s will still be there but you won’t.That’s my point that projections should not be used as a stick and here they were. So it drives me crazy when somebody is doing statistical models on projections.His point was you better not be hiring a super expensive VP of Sales and ramping a really expensive team when the market isn’t there.

      1. Tereza

        Phil you’re so right.You must’ve gone to one of those fancy-schmancy business schools.

      2. the slingster

        Yeah. Please note the fundamental points of these financial exercises are self evident.

    4. JLM

      Fabulous article. Starting my 4th decade in business, I can tell you with certainty that the “new market” curve has always been the rule rather than the exception. Inexperienced investors are as much the problem as inexperienced entrepreneurs. Every generation thinks it has discovered sex, but it is not so. So too, the pathway to riches — it takes twice as long and costs twice as much but if you hang in there, it’s all gonna be OK.

    5. fredwilson

      He’s right. A revenue plan is too tight, like a noose. A revenue model is not though. It doesn’t even have to be implemented. But you need to have some idea of where you are headed

  16. Scott Carleton

    I’m very happy your covering this Fred. When I first attempted to construct a basic pro forma I had very little to go on. I realized that the key was to make assumptions that were based somewhat on reality and just keep moving through numbers. The important part is to go back and continuously check your assumptions to make sure they don’t get lost in the noise.I’m very glad you’re going to be using real examples. Most help on the web for this sort of thing is too generic and rarely helps the individual startup. Showing a real company’s forecast and how they reached each part of it will be fantastic as a learning tool.

  17. paramendra

    Return Path’s mention takes this blog post to a whole different level.

  18. ShanaC

    First off:Thank you return path for your help.Secondly-Difference between bottom up and top down? (in terms of examples?)

    1. Gorilla44

      Top down forecasting – “potential market of 1,000 companies. We will get an initial market share of 0.5% and then grow that to 2% and then 10%.” This is a pretty weak way of forecasting.Bottom up – “we are selling product Y to companies. We have 4 sales people. They can call on 25 leads per week. They estimate 2 sales from those 25 leads based on prior experience selling similar product X. Initial Sales = 8/week. That can grow to 15 sales per week after the first year by adding 1 salesperson and getting data from the initial customers that will show that we save them $z/installation.”Bottom up is way better and shows much more analysis and planning.

      1. Tereza

        Lovin’ that example, Gorilla44.Top-down is a quaint starting point, to just check if you’re remotely in the ballpark of a business that can go anywhere. So if it’s not, pencils down. You give up or come up with a new or expanded market.Beyond that utility it’s really just mental masturbation and not actionable or meaningful.

        1. Fernando Gutierrez

          Top-down can also be useful to check your bottom-up numbers. To keep going with the previous example, maybe you keep adding sales people and growing, but maybe your market is not big enough to let you grow so much. Or maybe when you achieve some market share your conversions are gonna be fewer. It can put your bottom-up numbers in perspective.

          1. Tereza

            That’s totally true Fernando.I wasn’t thinking future iterations but you are spot on.

        2. RichardF

          love that last line Tereza

      2. ShanaC

        Thank you

      3. fredwilson

        nicely put Gorilla. that’s exactly right

  19. Gorilla44

    Projecting expenses is not that difficult, especially 1-2 years out. I run a biomedical startup with no revenues (and none coming soon) so my forecasts are 100% expenses. The numbers are based on experience and actual quotes from vendors – pretty easy.The little exposure that I’ve had to revenue forecasting has shown me that that is damn hard.I did run across a good slide deck on the subject that I recommend:

    1. fredwilson

      yup. expenses are easier than revenues. you have way more control over your expenses.

  20. Tereza

    This is a perfect structure — varying stages, and put into the context of a real live company! Clever.The advance planning you and the Return Path guys put into this will be to all of our benefit. It’s always of limited utility when left in the hypothetical.Sorry that I sound like a suck-up here, but I’m really looking forward to a rich discussion.

    1. Dave Pinsen

      I don’t think there has ever been a penalty here for sounding like a suck-up.

      1. Tereza

        LOL. Pucker up, Buttercup!:-+

  21. akharris

    Fred, I’m looking forward to this set. It’s a great counterpoint to the past post about future cash flows, and should be a great view into the other side for those of us used to running valuation models on companies. Since something like a DCF is, so critically, a “garbage in/garbage out” enterprise, I’m hoping this series will help give me a sense for what good projections look like, how to spot the garbage, and how to avoid it myself.

  22. JLM

    I am afraid to start typing this post because I have to go somewhere in 3 hours. LOLAs a numbers tyrant (engineeer, MBA finance), I can only tell you that numbers are very, very, very important but they are also as easy as pie to understand and create.First, understand that every business operates within a strict arithmetic algorithm which is likely driven by the 2 largest revenue accounts and the 7 largest expense accounts. Our job is to discover and tame that algorithm. I have paid a bit of tuition to learn that fact. May be just a bit different in your business but we have likely dealt w/ 85% of everything in those 9 line items. You have to discover that algorithm.Budgets have to be built from the ground up and they are to be treated with respect because they are real revenue and spending plans. Make folks adhere to them — particularly the salesmen who immediately want to lower the price.It is like swimming and diving — learn to swim BEFORE you learn to dive. It is generally a less stressful approach.Budgets must be updated whenever any critical element is shown to be wrong and at least every 90 days. Budgets are discipline.Forecasts on the other hand, are like those popular but mildly promiscuous girls your Mom warned you about. Never marry the forecast until it becomes a real budget but there is absolutely nothing wrong with exploring every facet as long as you realize it is all in fun.They can change quickly and they anticipate good times and rarely do they anticipate bad times well.When you update your budgets on a quarterly basis, take a look at your forecasts and see if they are still reasonable. Make a range of forecasts — a family of curves really — but resist the temptation to allow their nomenclature to suggest that one is better than the other.Graph your forecasts as it is illuminating to see how an early error throws off the entire family of curves in the later years.In the forecasting business, make sure you speak the tale — legend of the campfire type stuff — which supports them at the same time you are looking at the numbers. See if the tale is plausible and if it is something you have actually done before. Test it by your own demonstrated reality.Practice develops your skills and don’t fret that you miss everything, learn from it. When you can, hire a great CFO who loves this kind of stuff and work him like a dog.

    1. Tereza

      Another perfect 10.

    2. fredwilson

      you and i use the word forecast differently JLM and i don’t want to get hung up on words because it is what they mean that is me a forecast follows from the budget and is an attempt to determine what is actually going to happen when the business starts to trend away from budget.i don’t believe in rebudgeting intra year unless the company is way off

  23. rickfield

    FredYou are making these succinct posts a must read. My company makes pickles andone simple point I always say about business is this: we make pickles, but we could be making umbrellas, skateboards or mascara and the salient issues would remain the same. Worth remembering given the AVC tech focus.

    1. Tereza

      Rick please send us a link about your pickles.A good pickle is hard to find.

      1. rickfield

        You can check out Rick’s Picks at’s a video link, too: http://blog.rickspicksnyc.c

        1. Tereza

          Rick perhaps you can help me solve my perennial pickle problem.And forgive me, AVC folks, for distracting you away from Projections, Budgeting and Forecasting, for something that’s more appropriately a Gotham Gal conversation.Here’s the deal. I am very picky about pickles. There is a specific sweet/sour balance which Czech/Polish pickles typically have but is so hard to get in this country. They’re always either too sweet or too sour.I have some Polish and Hungarian delis I use as my standby. But even there it’s hit or miss, as the texture of the imported stuff is often rather….uh…droopy. (sorry guys)Summer’s around the corner and picnics are coming. But potato salad is pointless unless it contains the perfect pickle.Help me, Pickle Rick! What’s a girl like me to do?

          1. rickfield

            Tereza:You would find our Kool Gherks in keeping with your flavor profile preferences!BestRick

          2. Tereza

            awesome, will do!

          3. fredwilson

            terezathe gotham gal and i are investors in ricks picks. gotham met rick in the greenmarket in union square a few years back and took a liking to him and his pickles right away.all of rick’s pickles are fantastic but i am particularly partial to the phat beets, the mean beans, and the smokra. none of them are of the flavor profile you are looking for (for that get the Kool Gherks) but you should try them out.this is my favorite bundle…

          4. Tereza

            So funny — wouldn’tcha know I swung by the Union Square greenmarket yesterday 2pm looking for Rick’s table…in hope of grabbing a few jars and forgoing the $9 shipping!Alas they were not there but I’ll do a Wednesday.

  24. alphaG77

    Large tech companies generally use an Activity Based Costing approach for budgeting & forecasting. This allows them to rollup and benchmark the true cost of say doing a software iteration, fixing bugs, doing sales calls in a way that apportions overhead fairly to each. Unfortunately, setting up a robust accounting system, one that integrates with the significant day to day activities and operations, seems like too much overhead to undertake for most startups.However, it seems to me that it’s worth taking up at least in part early on, because it’s a much costlier thing to integrate into the company further on down the road. Anybody have a more experienced based opinion on that?

    1. fredwilson

      i have not seen any of the companies i am involved in do that sort of thing.

      1. alphaG77

        Not surprised – the higher the level of accuracy, the higher the cost of collecting the data!However, I actually think it’s a great approach to thinking about overhead, “indirect costs”. Probably more applicable in an e-commerce setting where a few points of margin can make a huge difference and where it’s critical to understand which activities and/or customers are most profitable.Take Etsy for example, could they tell you today which accounts are their most profitable relative to time spent serving them, including the cost of say writing community post, and broken out by account longevity? I don’t know, perhaps they have instrumented a system to facilitate answering those kinds of questions, but my guess is that it was a lot more stressful for them to do once they were pulling $50 – 100mm in revenue versus when they were pulling in $500k – $10mm.

  25. bizz help

    i don’t think present value of cash flows and mere projections are the real value of evaluating businesses because we can never take into a lot of other important issues like turnover, customer response, r &d, financial situation of the company and many other things into account

    1. fredwilson

      well ideally you can take all of those into account when you do the projections of cash flows

  26. JJ Freitag

    fred- one of my favorite tools for helping to report on and tune the forecast is the waterfall chart. I love the look back mechanism it provides to help companies (and board members) get better at forecasting by tracking the original budget and then revised forecast each period (typically monthly). I have found that both mgmt and board members get better at setting expectations over by tracking this rolling forecast and reviewing it monthly and it can easily be populated from the P&L. Happy to send an excel file to anyone who would like to see an example of some that I have used successfully.

    1. fredwilson

      you need feedback loops to refine and tighten up the projection, budget, forecast process. a waterfall is one of the best.

  27. fredwilson

    agreedi plan to get into all of that in the coming weeks

  28. Jared McKiernan

    one of the traps in building forward-looking projections is that many people do not realize some of the assumptions their models take for granted, and in which scenarios the model could fall apart.Thinking about long-term results as a range of possible scenarios and outcomes, and talking through “Why?” for any givens in the model can easily turn up problems inherent in choosing underlying metrics which are mostly statistical noise in the way of what little real predictive power in such a model. Total market size is definitely not good enough, and also anything based on comparing to a rate chosen from a prior start-up’s experience is probably way more likely to do harm than good. Another one that seems common in long-term forecasts are absurdly unsustainable compound growth rates.I think this blog’s collective wisdom would probably result in solid forecasts; along similar lines, it would be great if the archaic gambling laws on prediction markets are changed to stop standing in the way of innovating more accurate information.

  29. eric_bingham

    Detailed underlying metrics, backstopped by reality, are essential. Biggest mistake I’ve seen in looking at plans (and made myself) is having multiple growth/revenue variables increasing across the 24-36 mo. plan. That creates an unrealistic compounding effect in the numbers that’s not initially apparent. Each of the increasing assumptions themselves are defensible and even realistic. But in combination create a plan that’s not – or at least becomes largely unattainable as not everything in business moves in such a straight monthly line.

  30. fredwilson

    i find bottom up approaches much better than top downfigure out how many customers you can likely serve in the next year or twoand what that would be worth to you

  31. Mark Essel

    Thanks Alex, this is one of those nuances that’s great to know (start with now and look at your edges, vs assuming a piece of a huge pie).

  32. Greg Gentschev

    In my mind, there are two separate exercises to think about. One, what’s the market size? Like Fred mentioned, bottom-up market sizing is a great reality check, but I like to look at both that and a top-down sizing to see how well they match up. I’ve written a couple of posts about that here:…. I know, blatant plug, but I think it’s actually useful.Two, how much of that market can the company actually get? I think that’s what the upcoming projections and budgets posts will focus on, but that’s where it’s really critical to focus on the company’s existing distribution channels, conversion funnel, and revenue per customer. Dave McClure’s written some good stuff on that, too.

  33. Dan Ramsden

    But like I said, not in terms of proximity to financial targets, but in terms of the very business model that materialized. As this could change the structure of the forecast completely.

  34. Dave Pinsen

    Is this analogous to estimating a public company’s forward earnings and applying an appropriate multiple to it? I know some investors like to do a DCF analysis going out a bunch of years, but I think with tiny public companies (some with fewer employees than some of your portfolio companies) there often isn’t much clarity beyond a year or so.

  35. paramendra

    Don’t apologize for linking to your blog post any more than you would for taking up space to comment here. It is still you talking both times. A link is an option to click over or not.

  36. fredwilson

    There’s a ton of great stuff out there on this topic. Thanks for adding the links. Shameless plugs to relevant content are called recommende links around here

  37. fredwilson

    Yes. A multiple of earnings is a “poor man’s” DCF

  38. fredwilson

    Yup. Garbage in garbage out

  39. Tereza

    Garbage in, garbage out is definitely true for financials and data analytics.But — to segue to today’s post — it is certainly not the analog for your hiring process. In fact the opposite.No doubt many treasures in that group. I for one would love to sift through your castoffs (I have my eye out for a mini-me).Fred, when you do your post-process ‘honoring of the candidates’ (live or virtual), would you consider involving promising local startups who might have work (paid or unpaid) for some of these people?

  40. fredwilson

    using a consistent month/month growth number over 24-36 months is most definitely a recipe for problems. you just cant’ grow at the same rate for long. the law of large numbers works against you