Discovery Driven Planning Case Study
click for presentation: Discovery Driven Planning - Annotated
Wharton's Ian MacMillan and Columbia's Rita Gunther McGrath, articulated a system they term "discovery-driven planning" in their August 2000 book, The Entrepreneurial Mindset. "[D]iscovery-driven planning," they wrote in an earlier article, "acknowledges that at the start of a new venture, little is known and much is assumed." This they contrast with conventional, "platform-based" planning, in which "assumptions underlying a plan are treated as facts -- givens to be baked into the plan -- rather than as best-guess estimates to be tested and questioned."
For launching a new venture, they advocate drawing up a checklist of all key assumptions, a timetable for proving each one, and what they call a "reverse income statement" -- a list of threshold economic criteria that must be met if the project is to go forward.
Their system is meant to apply to any entrepreneurial venture. Developing a business plan for an existing line-of-business (LOB) is quite different than planning for a new initiative. To plan for an existing LOB, one could extrapolate future results from past experiences. But for a new initiative – whether a new business, new market or new product – extrapolation rarely works, because, the predictions for a new initiative are based upon assumptions and not past knowledge. So what is the best planning approach for a new initiative?
Discovery-Driven Planning is a tool that converts startup assumptions into knowledge that grounds the planning for a new initiative in reality.
The tool imposes planning discipline through four documents.
- ‘Key Assumptions Checklist’ identifies the business hurdles and assumptions for the initiative.
- ‘Reverse Income Statement’ models the business economics.
- ‘Pro-Forma Operations Specs’ defines operations needed to run the business.
- ‘Milestone Planning Chart’ specifies when the assumption needs to be tested.
A simple case shows how the process works. Consider a proverbial widget making company, which is contemplating entering a new market with a new product. Its return on investment and other directives call for a profit of $400,000 for the first year from the new initiatives. Some of the other key assumptions:
Reverse Income Statement
Be specific about what you expect from the business using the reverse-income statement as a tool. (e.g., required profits = necessary revenues minus allowable costs.) Don't start with the costs and hope the profits will flow. Figure out what profit will make this venture worthwhile, then work backward to figure out what it will take to get there.
The initial Reverse Income Statement. Discovery-Driven Planning starts with the required profit and works up to determine how much revenue is needed and the cost allowed to generate that revenue. In the above example, the required revenue would be $4,000,000 (a 2.5% market share), allowing a total cost base of $3,600,000.
Pro-Forma Operations Specs
Next, develop the Pro-Forma Operations Specs.
Milestone Planning Chart
Finally, develop a Milestone Planning Chart. Plan milestones that are targeted events and goals not arbitrarily chosen dates. Focus on converting assumptions to knowledge. At each milestone, revisit the assumptions and business model, learn from it and adapt accordingly?
This entire process is looped and there is continuous feedback from one stage to next. Such a planning tool allows for a rigorous testing of the assumptions that necessary for a new initiative throughout the initiative. It has strong potential to reduce the uncertainty in planning for a new business. It enforces a planning discipline that creates checkpoints before significant commitments are made that could help the management determine whether it is prudent to go forward or not.
Discovery-driven planning is a powerful tool for any significant strategic undertaking that is fraught with uncertainty - new product or market ventures, technology development, joint ventures, M&A programs, strategic alliances, even major systems redevelopment. Discovery-Driven Planning forces managers to articulate what they don't know, and it forces a discipline for learning. As a planning tool, it thus raises the visibility of the make-or-break uncertainties common to new ventures and helps senior managers address them at the lowest possible cost.
Focusing on the adaptive execution of a business, Discovery-Driven Planning is also useful in managing existing business lines as it helps managers focus on key assumptions and adapt to changing external environments.
Implementing a DDP involves every aspect of operating and growing a company (including identifying, assessing, developing, and launching new products and services), but is rooted in extensive financial analyses and is centered on significantly growing revenues and reducing costs.
Discovery driven planning is a plan to learn, not to show that you had all the answers when you wrote the plan. The technique requires the interaction of five processes, working together. These processes are: 1) determining the frame (objectives) at the level of a project; 2) establishing competitive and market benchmarks; 3) defining operating specifications; 4) documenting assumptions; and 5) establishing key milestones. In uncertain environments, conventional planning makes no sense. Instead, plan with discipline to the next major milestone, then pause and re-plan as new information becomes available.
So we've been talking about how assumptions underlie any startup business.
And how figuring out what those assumptions are and planning around them,
planning to learn about them, is critical to making your startup work.
I'm going to give you a technique that was developed here at Wharton called discovery
driven planning that'll give you some ideas about how to think about formally
planning your business.
This was originally developed by Ian MacMillian and Rita McGrath, and
it's a really interesting way of building a toy model.
What I mean by toy model is this is not trying to reflect the real world in a deep
way, but it instead creates a steady state model of your business that lets you
play with various assumptions and figure out what's working and what isn't.
And what's nice about this is unlike a lot of other models that are financial,
it takes into account the fact that you don't know a lot of answers.
So, the variation I'm going to teach you about discovery driven planning has
And before I explain the five steps,
I want to explain how we're going to go through this, so
I'm going to give you a little bit of a case that's a made up case.
We'll walk through how you would do discovery driven planning for
that case, and then available as part of this course,
you will find a spreadsheet that will actually let you play with this model.
So, the model I'm telling you, we'll built a spreadsheet for it.
And then the second video lecture, on discovery driven planning will actually
ask you some questions about that case, and will dive into more detail.
So, it's a complicated topic, it's much easier when you do this hands on.
So use this as the introduction, then play with the actual spreadsheet, and
then the second video will give you a sense of how it all comes together.
So, it's much more intuitive once you have your hands on this.
But the five steps you have to worry about are starting with a goal, mapping of
the operational steps for the enterprise, build a reverse income statement,
benchmark your assumptions, and match your milestones to assumptions.
So, what does this all mean?
Before I go into them in detail, let me give you an example.
So, let's take the example of your cousin, Jon.
Jon and his wife Joanne are carpenters.
And they've been working as cabinet workers for a long time.
And they have been working in a bunch of different fancy office buildings,
and they've realized that the kind of desks, especially standing desks that
are available to people working in these offices are pretty ugly, and
not very functional.
So, they've come up with a way of making a really beautiful standing desk.
We will doing creative work.
And because they have modern carpentry machines that they've been using
to do their contract carpentry work,
they actually think that they can do this much more cheaply, and much more nicely,
and much more ergonomically than any standing desk that's out there right now.
So, that's their idea.
The company is called Better Desk, again, this is fictional.
And Jon and Joanne have asked you to think about how they should plan their business.
And we're going to build a discovery driven plan for Jon and Joann.
So, the first step is to start with your goal.
And this is part of what makes a discovery driven planning interesting is,
like most entrepreneurs, you're starting with your endpoint.
What do you want to achieve?
What ultimately do you want to get out of launching a new venture?
And in discovery driven planning, you think about that accomplishment first.
And now, since we're building a financial model, we need some numeric answer here.
So that could be that you want to achieve is a particular level of yearly profit,
Or that you have a particular price point, when you want to make a $100 drone
that follows you around while you do extreme skiing and takes videos of you.
Whatever outcome you want, that's where you start.
In the case of Jon and Joanne, we've talked to them,
and it turns out that what they'd want to do to launch their business,
they're excited by it, they think it's interesting,
they really want the freedom associated with launching a new business.
But they also want earn more money than they're earning today.
So, they want to take on 15% more income than they do today,
they're each making $60,000 in our example.
So that means that their business needs to generate $138,000 in profit,
15% more than a 120,000 in order for them to bother launching it in the first place,
so that's our goal.
We need a $138,000 in profit for Better Desk.
So what's the next step?
We now need to think about the steps we actually need to go for
in our enterprise to make it work.
So, what are the things that we will need to do?
Now, if you remember back in the previous lecture we talked about assumptions,
this is your technology in operations set of assumptions.
What are the actual concrete steps your business needs to go through in order for
you to be successful.
So, what I recommend doing is working backwards from
delivering your product to your customer,
and thinking about the operational steps that you need to go through.
So in this case, again, I'm not a desk expert, so I made this up.
And for those of you watching this video who are desk experts, I apologize for
everything I've gotten wrong about the desk manufacturing process, but
this is designed to be an example.
So we start with the end point, which is we sell our product to customers.
So before that, we need to ship it to them and deal with returns.
Before that, we need to actually manufacture these things, so
people can use them.
We need to go through some sort of ordering process,
where we book orders and set them up for this.
And we also need to have made people aware of this before we do anything else, so
we have to do sales and marketing.
So these the broad steps of your enterprise.
Now, we then go into detail and think about the operational specifications,
the individual tasks and metrics that matter in each step of this business.
So in the case of manufacturing,
here's my example of the sets of manufacturing steps that I thought about.
So, there's some raw material for desk,
there's some delivery cost for that raw material that results in material cost.
So that's one bucket that we have to worry about.
Another is that the production line itself will turn out a certain number of desks
per day, and how many days per year are they going to be spending doing this?
How many production lines are we going to produce?
And then figure out the equipment cost for
the production line, any depreciation rate, and then the staffing.
How many manufacturing staff do we need?
How much we're paying them?
These are the direct operational specifications for the business, and
you'll think about getting all of those on paper.
So you've broken down each step of your business, and
we can do the same thing, you can imagine for sales and marketing.
Thinking about how much we're spending on doing advertising and marketing,
who's going to do this, how many salespeople we need, how many orders
they book per sales call, how many sales calls could they make per day.
So we're getting very detailed into the operational specifications for
Now, please keep in mind that we won't have all the answers at this stage.
So, you're going to do your best job possible, and
even the process of going through this has benefits.
We know that businesses that spend time on planning are more successful
when they launch than those that have not spent this time.
So, there's benefit to even going through this,
even if you don't have all the right answers.
You now go from having all of the operations specifications of your business
setup to building what we call a reverse income statement.
So unlike a regular income statement,
where you're generating a based on what you're spending and what you're taking in.
In this case, we start from our goal, and so
if you look at this sheet, you'll be able to download this again and
play with these numbers, we have this available for you as a spreadsheet.
And this is for the better desk case, but
you'll notice that some of the numbers are in green.
Those that are in green are assumptions,
these are numbers we're estimating based on some sort of characteristic.
And the numbers of regular text are calculated from this.
So we're starting with the current salary, you'll notice at the very top, Mr.
and Mrs. Smith are currently making $120,000 a year.
We then start with our first assumption, which they've told us they want to
earn 15% more, that means the required profit for Digital Desk is $138,000.
And now if we skip down to the manufacturing costs,
we'll see that the raw material per desk, we're guessing a number.
The delivery cost that we just discussed before, we're guessing a delivery cost.
We're figuring out,
we're assuming how many desk produce per day in this production line.
We're assuming how many production days that are per year.
So we're filling in numbers, those operation steps, and
linking them together in a reasonable way.
So, all these costs are being summed up,
and then being subtracted from revenues to see whether or not we're still over that
required $138,000 goal that Jon and Joanne have.
You now have a list of assumptions in your business, and
the goal then is to figure out what are the best possible numbers that I might
have that may match these assumptions?
So, we don't know the raw material per desk.
So, what do we do to get that answer?
So, we draw on our own personal experience, but
more importantly we call competition,
we look for numbers in the press, and we try and get the best number we can.
So here I've created some fictional sources for all this data.
So, in figuring out whether the raw material per desk, we have called
our potential part supplier Hardware.com and have them price out the numbers.
So, we have a pretty good estimate that's $120 in raw material,
and they also told us it'll be $10 shipping cost.
When we start thinking, even things like the desk per day per production line,
we're now going to our equipment supplier and saying,
how many desks can we produce in this equipment?
And they're giving us that information, that's what's being plugged in there.
But then, when we start looking at things like,
further down, the manufacturing staff per production line.
We have that it's going to take six people per line.
In this case, we're benchmarking based of on our best estimates of competitors.
So, we've spoken to people who know the competition and
they're telling us that it's six people per production line.
And they're telling us that the salary is $35,000.
Now, those assumptions might be more vague for us.
We don't know the answer as well,
as opposed to the calls we've had with the hardware manufacturer.
Who have been pretty clear about how much we're going to have to spend.
So, each assumption that goes into our chart, every number here,
has a source for it.
It's either calculated from other numbers, or it comes from some outside source.
These are your assumptions.
So you put those together in one chart, and
here we list a whole range of potential assumptions, values, and sources.
And again, you can play with all these numbers in the spreadsheet.
So step 4, you're benchmarking your assumptions.
You're figuring out what numbers you can use to fill in the blanks for
your assumptions, for those green numbers, until you have better answers for
Step 5 is where this discovery driven reverse income statement starts to
become a planning document, and this is about matching milestones to assumptions.
So, you can't just have assumptions in your discovery driven plan,
you need to actually test those assumptions.
And you want to start, just like we talked about in our prior session,
with the highest uncertainty, highest impact assumptions.
Those are your key assumptions.
So, you're going to figure out what your key assumptions are by looking at
the model that you have, and you'll start by discussing this with your team.
Do we know an answer to this problem or not?
Is it high uncertainty or not?
You could do sensitivity analysis.
The nice thing about discover driven planning is that you have a document
that now has all the numbers in your business all tied together.
So, if you move those numbers in a range,
I change a number from say, six employees per production to seven.
What does that do to my underlying cost structure,
and what does that do to my profit margin?
If I changed it to five what does that do?
So, I can actually get numeric answers by changing those numbers.
And I can also look at the list of assumptions.
When you think about Jon and Joanne, they're carpenters.
So, I probably would feel very confident about their numbers,
about production cost because they know a lot about making wood objects.
But I would be less confident potentially in their sources and
uses for their pricing because they never built and sold their own desk before.
So also comes from the sources, or
using to fill in those blanks in the discovery driven plan.
You're going to match these assumptions to milestones.
Milestones are moments in your business when you're getting real world feedback
and information based on your real product, or service, or venture.
So, we'll talk a bunch of methods later on in this class session, we'll talk about
customer interviews, we'll talk about prototyping, we'll talk about surveying.
So, you're actually going to be getting data from the real world,
and that data will provide milestones on your business.
Ways of updating your assumptions, and learning about how the real world operates
in a way that lets you modify your business.
So, just like key assumptions are the ones that have the highest levels of
uncertainty and the highest levels of concern,
key milestones are those milestones that test the most assumptions.
So in the case of Jon and Joanne, we've created a bunch of potential milestones,
again, I don't know the carpentry and desk market that well, so this is hypothetical.
But they might start with a market study where they're going to gather as much
market data as possible, they'll purchase market research on desks, and
they'll do some surveys of customers.
So that's an easy milestone before they even start production.
The second thing they might want to do is a trial production batch.
So they might build a few desks, and see whether or
not their techniques are correct.
If you think about it, these milestones test different things.
The market study is about testing the customer value proposition.
Do people want this desk?
How do they buy a desk?
What does the market look like?
And the production batch is about test technology to operations assumptions.
How much does it cost to make a desk?
How long does it take?
So, each of this milestones are the steps in their business,
each milestones will be testing the underlined assumptions.
So, you actually just create a milestones assumptions chart.
All the assumptions that we've already talked about are listed on
the left-hand side.
So, the direct sale price for desk, the raw material cost per desk,
things that were already in our model.
And then on the top, the columns, are the milestones in the business so
the market study, the trial production batch.
And what you're going to do is match the suit together by simply putting an x,
where a particular milestone test particular assumption.
So, assumptions might be tested multiple times, for example, we might test the raw
material cost per desk, when we do our trial production batch.
And then we'll get more information when we do our mock-up, or
even more when we start production.
So, you will narrow down the ranges of
those assumptions as you get more information at each milestone.
So, this is a process of learning real world information to change your
assumptions, and to narrow those assumptions down to get better information
and build your model more accurately as you go.
Then you're going to I think about how you order those milestones in ways that will
minimize your cash burn while you're learning.
So, one of the cheapest milestones you could do,
a survey is much less expensive than quilting a job and launching a business.
And how do we learn as much as we can from those early milestones early on, and
then you're going to actually use this milestones to help shape your business.
During the on-going operations of your venture, as you go along,
you can keep going back to that milestone assumption chart, and
castly check the validity of your underlined assumptions.
And you'll update your model with those numbers, as you update that model you're
going to get a sense of where the problem areas your business on financially.
Whether you'll be able to reach your goals, and
you'll get advanced warning of potential concerns in your planning, and
that'll let you revise your plan as you go along.