Worksheet For Developing Your Logic Model Worksheet For Developing Your Logic Model: Fill & Download for Free

Check out the incredible wealth of free resources (below) available on good spreadsheet practice developed by some very dedicated and brainy people with YEARS of experience designing and auditing Excel models as their business. It is not much fun reviewing or using someone else's crappy Excel model, or being expected to provide critical business decision support on the basis of results generated -- these guys have seen it all!Here's some tips:(but I do encourage you to look carefully at the resources below -- no affiliation with me -- because it is a big topic and few get it right).Be pretty. This means clear structure and formatting (inputs - calculations - outputs - checks), consistent style, easy navigation, useful tables and charts, dashboard and one pager summary printouts etc. You will be amazed how much kudos you receive and confidence others have in your model, despite knowing nothing about the underlying calculations.Be left-right consistent (for calculations on a time-base). The "intellectual property" in your calculation worksheets should reside solely in the first column of your calculations. Theoretically you should be able to delete your entire model except the first column, copy left-right, and regenerate the entire model and get the same results.Start simple. Use the principles of agile development - get your basic model up and running, then enhance it later based on feedback. Many times additional complexity does not enhance the business purpose of your model, and in fact can be a negative. Make milestone dates all start on the 1st of the month or quarter for example - good enough for a first pass. Dealing with partial period calculations is one of the areas most prone to errors.Don't double-up. Do inputs and calculations once, and link to it for other uses. Another area prone to errors is when two inputs are not mutually exclusive - for example, a capex profile and a second input for construction length. Use one input to calculate the other, or at least include an audit flag to alert someone that the two inputs are inconsistent.Test it. Tame your ego - your model WILL have errors. There will undoubtedly be logic conditions based on different combinations of inputs and results that you NEVER anticipated in your formulas. Use a systematic method to check your model before it is released in the wild. I use Monte Carlo simulation. Set all your model inputs to vary over the widest reasonable range, and see if your accounts still balance or your results make sense after 1,000,000 random simulations - I can almost guarantee it wont (if you use calculation iteration for circularity, this is where you see it all go wrong!). Despite 15 years trying, I have never attained the elusive goal of getting no errors first time, and I never expect to either. After systematic testing and making corrections, at least I have the confidence that my models are as near as possible to bullet-proof when I deliver them.Corality Financial GroupFinancial Modelling Handbook - The first collaborative, crowdsourced financial modelling guideBest Practice Modelling StandardsSpreadsheet Standards Review Board

Generally, unless you have real evidence that your business model is valid (i.e. at least initial sales), there is no logic in creating financial projections. The reason is that it will all be based on assumptions, which are usually very far from reality.If you don't have real evidence that validates your business model, then you only need to focus on a detailed budget. Serious investors anyhow won't "buy" a detailed financial projection at this stage.Now getting into more details, your question includes two very different scenarios. The first one is pitching to VCs for a seed round. The second one is pitching to VCs for a Round A.In the first scenario (i.e. pitching for a seed round), you'd probably be better focusing on angel investors rather than VCs. VCs tend to invest only when the business model has been validated, and this usually only happens after seed funding (and kudos if you already are without any funding). Therefore, focus on creating a detailed budget, since angel investors would want to know what's your burn rate (i.e. how fast you're going to burn their money).In the second scenario (i.e. pitching for Round A money), you must have already validated your business model. Otherwise, you have almost no chance getting any money. Assuming you already have strong evidence that validates your business model, then it's about time to build a serious business plan.At this stage, you should already know your Customer Acquisition Cost and the Customer Life Time Value. You should also have good matrices that can help you project your future growth rate. With all this information in hand, you can quite easily project into the future and convince investors that your projection is realistic.And one last tip. The most important worksheet in your financial projection is the Assumptions one. You should have a detailed and separate assumptions worksheet, that would allow investors to easily see the logic behind the numbers, and play with the assumptions in order to test different scenarios.

I get it. You already know R, or Python, or some other language. And you are incensed that you can’t use that language to build Office macros. Well, the fact is that Office macros can be written out of the box in either VBA or in Office-js (based on JavaScript). You can also use the Office interop with the languages in Visual Studio. In addition, there is limited support (from non-Microsoft sources) for other languages such as R and Python.Any language that you don’t know and are unwilling to learn is going to seem ugly and objectionable. That ugliness isn’t an inherent property of the language—but it might be a reflection.When you program macros in any language, 90% of the complexity derives from the objects, methods and properties of the application object model. So are you saying that the Excel object model is ugly and objectionable? I would much rather leverage the VBA equivalent of the worksheet user interface than program a PivotTable or sort routine from scratch.When evaluating languages for scientific and engineering software, I place high priority on code readability. Most of that code will be in user-defined functions, and relies on a handful of language features inherited from Visual Basic. Your non-programming peers and boss should be able to read your code and understand what it is doing. They aren’t required to write that code, but they ought to be able to figure out its building blocks, trace the logic, and perhaps troubleshoot a bug. VBA is pretty good in this “human readability” thing.Languages should also be easy for a neophyte to pick up. I have hired a number of college interns for summer jobs, given them programming tasks, and found them ready to start in VBA after an hour of training. They won’t be producing elegant code, but it will work and be useful from the outset. I find VBA’s quick learning curve a very attractive feature.You might also complain about the lack of open-source libraries available to VBA programmers. Python and R have a rich assortment of open-source libraries because those languages are favorites of the academic community. VBA receives much less love. But VBA can use libraries if you set a reference in the Tools…References menu item. That’s how you get access to regular expressions and dictionaries. VBA can also call code from add-ins produced by third-party developers.Rather than complain VBA is ugly and objectionable, I think it ought to be the language taught in computer programming service classes. Almost every computer in the business world is equipped with Microsoft Office, and all of those users have everything they need to improve their productivity by writing VBA code. I call that beautiful.