Santiago Bilinkis l Risk and Reward - Trying to live a life out of the comfort zone

Considering that there are many people who asked to see an example of what I have built, instead of sending it to each person I am going to upload it as a post (it’s too long to be a comment).

It’s a little boring to read, so I recommend that only those that are dying to have an example of what I talked about in the last post, read this. In saying this, I feel liberated to post something long and difficult to follow, and it’s only for the really motivated who want to know more about this. If you get bored, don’t complain! Before you dismiss me, think of the work I did to make this “complexity”.

To get things straight, this is not a complete list. This is a more detailed explanation of how to build a business model in Excel, using Officenet as an example. I have the list that we once used to plan ON before launching but without explaining it would result in something incomprehensible, I think.

Let’s see… At least before having any employees, we planned how Officenet was going to do month by month for the first 24 months, which seemed more like an one of Horangel’s exercises in divination rather than an entrepreneur. For example, the sales… How do we know how much we are going to sell?!?  But what you need to try to do here is not to plan the “large” and uncontrollable variables, like the sales but to break them down into to their smaller components.

We thought:  What do Officenet’s sales depend on?  Alright, the total sales is the quantity of customers that buy from us in a month, multiplied by their average consumption.

And how many customers buy from us in a month?  That depends on the quantity of customers I have at the start of a month (the first month is zero after that no), the quantity of new ones we get and the quantity of old ones we lose.

And what does the quantity of new customers we get depend on?  With that, we simplify it a little, Officenet’s client search is through direct marketing (delivery of merchandise through mail),  depending on the percentage of people that respond to us and open an account through the catalogue we send them, multiplying that by the quantity of goods I sent. This last variable is already a good start because the decision of how much merchandise to send was our own decision.  If we had wanted we could have sent 10, or if we wanted 10,000.

The list of customers for the first month is the quantity of merchandise that I send (that I already know for sure because I decided it - going along with the last post that we colored green)  multiplied by the effectively rate of that product (data that I don’t know but that is based on experiences in direct marketing campaigns and is much more simple to estimate with true precision than something that is so general such as the sales - yellow variable). I don’t lose anything because I didn’t have any from before.  Multiplying that quantity by the average consumption ( that is unknown to me and I estimate crudely - red variable)  I have the sales of the first month.  The second is done in a similar fashion.  I take the quantity of customers from the past month, then I take out a percentage of loss (another estimated without exactness- red), I sum up the new ones according to the amount of deliveries I do and that amount I multiply that by the consumption.  And so on…

This same criteria of breaking down a “large” variable like the sales into small components, some better known than others, it must be applied to break down each one of the variables in the business (margins, structure of areas of employment, etc).

How many inquiries am I going to have each month?  For that, you must divide the total sales by the average amount of inquiries (another red variable), that is not the same as consumption because some customers buy more than once per month.  Having that number of inquiries that I have, make, and send; I can start to plan the areas of operations.

How many people am I going to need in the warehouse?  How many trucks do I need to make the deliveries?  It is simple to estimate.  We split up our thinking to how many inquiries can a person fill in an hour (yellow) and how many hours are my employees are going to work per month (green).  Knowing how many deliveries a truck makes per hour is the same as the number of trucks.  You have to do the same to calculate the size of everything in the area of operations.

But, be careful!  Not everything depends on the number of inquiries!  Some depend on things like sales, the amount of customers (the number of Account Executives), the total of employees (personnel at the IT help desk) or things that don’t depend on anything at all (amount of people in the area of Sales).

Having given dimensions to the areas of operations, we can begin to estimate the structure of costs. We added to the list of “possibles” how much we believed we should pay the people doing each job.  The amount could be set or could include some commissions that depend on some of the past variables.

The next variable is the crude marginal percentage (the difference between the price that they are going to pay for each product and the cost for me). In a business with miles of different products, each one with a distinct price and cost, is perhaps the most difficult variable, because to estimate it well would require knowing how many units I am going to sell of each one.  If I don’t have any idea, “how can i do it?”.  I look at companies like Staples, Office Depot and OfficeMax’s balances, that they quote in the stock exchange and whose balances are public information and I assume my margin will be equal (o x % lowest) to theirs.  Is this a good estimation?  No, but it it the only one that I have in this instance, for that it suits me fine and I color it red.

Now we are missing the application of this same principle to the monetary variable (typically, the work wages) in order to completely understand our first draft. To calculate, for example, the monthly evolution of the payment accounts, charges and inventory that I am going to need in the deposit.  I start to estimate the average period I am going to sell (yellow because I don’t know how much financing will exist) and multiply it by the monthly increments of the sales.  I apply the same to the other two.

At this time, don’t forget to calculate the taxes!!! This is an super common error to make, like forgetting the things such as the IVA or the taxes on Earnings…

Having done this, we have all the elements we need to build the chart. First,  we create a “tab” called “possibles” where all the possible variables go, coded their respective colors.  Then we go on to make other tabs to model Sales, Margin Operational Structure, Cash flow and eventually Balance.

It is fundamental that within all of the tabs, I save only the possibles that have formulas.  All parameters that have to be entered in by hand must be part of the Possibles. When a formula needs information,  for example,  the amount for an average inquiry, you must read it here.  If the “possibles” change with time the “possibles” page cannot have only two columns (the name of the “possibles” and its worth) but various columns like months that permit us to alter the timetables of value.

After finishing making this, we should be able to modify the worth of any “possible” and with that change automatic repercussions in the whole model. Example: if you change the quantity of units of directs marketing that you send each month and that part grows at a faster pace and the sales drops quicker.  Then before contracting more people and trucks, I need more inventory and charge accounts, etc.

If the model works well, we keep doing two things.  One is to start working on eliminating the “reds” and the “yellows”.  The other is to make an sesitive analysis.

This second task is fundamental, because sometimes there are variables the we have in red because it is a roundabout estimate.  But when we play with increasing and decreasing the value a little we realize that it has a very small impact in the results in our model.  If it is like this, it is really not worth your time to figure what the exact value that will make it “green”.

We will see that other variables that have an enormous impact (for example the response rate to the direct marketing campaigns, the margin, the average consumption of the customer, the rate of monthly clients lost, etc.) Those are the things you should concentrate on the most in preparing.  Brushing up the estimation can give merit to traveling to similar companies (like the margin, the average consumption or the rate of losses) or testing it out for real (send 1000 units to potential customers to see the response you get, even if you don’t have an actual business yet to give the service you are offering).  We aren’t finished with this process or present it to an investor without real studies behind these neuralgic points of our business,

If you are really bananas, you can play a little with a specific software program making simulations of Montecarlo.  Here are some examples.  I used one called @Risk but I think there is a free one made here in Argentina (I have never tried it).

When the final model is sophisticated and reflex well what the company should be, and when most of the possibles (including the most sensitive ones) are green, we’re ready to go to next phase: the process manual creation.

A final statement. Not all projects have the same complexity level to be modeled. I imagine that some of you are trying to start tech projects, so lots of the complex things to model about Officenet (trucks, stocks, etc) would not be relevant to your business. Maybe there are some others not covered in this post. The important thing, is to create a model to simulate the business behinf our idea, as best as possible.

I guess this clarifies what I wanted to say on the post about idea theft. After doing all this work, tell me if you still fear someone can steal your idea!