Calculate Your ROI, Make Two Envelopes...

Calculate Your ROI, Make Two Envelopes... David Liss February 4, 2002

Project Portfolio Management

There is a story about a man who, on starting a new job, goes to his office and finds nothing there except two envelopes, labeled "one" and "two," in a desk drawer. There is an attached instruction sheet that says, "Open letter one when you get into your first office crisis and letter two when you have your second." Sure enough time goes by and he gets into a big mess. He opens the first envelope and it says, "Blame everything on your predecessor." And so he does and all his problems magically disappear. Months go by and he gets into an even bigger mess than the first time. In a panic he remembers to open the second envelope. It says, "Make two envelopes..."

Ahem...what, you may ask, does this have to do with developing a return on investment or an ROI? Furthermore, what the heck is an ROI anyway? Regardless of anything else you read here, just remember the punch line to the story above as you read along.

Defining ROI
According to Michael Gonnerman (Michael@gonnerman.com), a Boston-based financial manager and contract CFO for high-tech companies, with over 35 years experience, ROI is a confusing and often misused term. "ROI is a measure of cash collected to cash spent adjusted over time. If you spend $1,000 and collected $1,200 the ROI is 20% for a single year. The longer the payback term, the lower the effective return is when considering the time value of money." Put in another, much more complicated way, according to standard financial dictionary definitions, ROI is the discounted present value of future cash flows generated by money spent.

When incurring any kind of expense, some kind of ROI method should be applied. There are no generalized models of what is a desired ROI for all businesses. A company can estimate ROI based on internal models and the validity of data. Everything has an ROI, the problem is figuring out what the right one is.

If you spend cash on something, a computer for example, then you expect to get a measurable return for the money spent. If the purchase is to be considered cost-effective, the cash coming in as a result of the purchase should exceed cash going out. Phil Rettew, a Senior Market Analyst with Merrill Lynch, explains ROI as what you get for what you spend for a described amount in percent on an outlay over a period of time.

Another consideration relative to ROI is the payback period, that is, when you spend $1,000 in one area, how long does it take to recover costs by increasing savings in other areas or by increasing revenue? How long would it take to recover a $100,000 investment? In the dot-com boom, the payback period typically was a multiyear consideration. Now investors are looking for an immediate return on investment.

ROI v. Cost-Benefit
By and large, the public sector in reality does not use ROI or payback considerations. Instead many public sector entities use cost-benefit analysis. There is a distinction between cost-benefit analysis and ROI. ROI relates the amount of the benefit to the amount spent while a cost-benefit analysis is a technique for comparatively assessing the (monetized) costs and benefits of an activity or project over a relevant time period.

Gonnerman went on to say, "You can't do ROI on fixed infrastructure costs, as infrastructure investments are not based on choice, and therefore ROI considerations are not applicable." However, for infrastructure-related items, how should you consider costs and benefits for a mandatory purchase? For example, you have to have a phone system. Which phone system do you pick? What are the functionalities and related costs of different manufacturers? An executive who indiscriminately allows any phone system to be used without first considering features, functionalities and internal company needs is, as the French say, a knucklehead.

The same kinds of considerations apply when an organization has to spend money to comply with legal requirements or organizational directives. Much financial expenditure, regardless of how it is dressed up, i splumbing considerations: getting an accountant, putting in a new general ledger structure.

If you can measure the cash benefit from the investment, in the form of increased collections or decreased expenses, then you can do an ROI calculation. Often it is easier measure cash going out, and more difficult to measure cash returning to the company. Few people, unless they are accountants, define ROI as stated above. Others (like me) think of ROI as a nebulous concept.

If you are trying to sell something, the customer often wants to see an ROI--so a salesperson will often be put in the position of "creating something" (think of the story above, make two envelopes).

Gonnerman also mentioned Economic Value Added or EVA as an additional ROI type consideration: At the entry level, this refers to the amount earned by an entity that is in excess of the cost of capital. If the cost is $50,000 and you earned $51,000, the difference between the cost and the amount earned is the EVA.

Figures Lie and Liars Figure
Joshua Greenbaum, Principal of Enterprise Applications Consulting, a Dale City, California, enterprise software consulting and analysis company, says, "It is a popular myth that business is governed by a series of rational processes used to arrive at rational decisions. You have to approach the business decision-making process as a pathological situation. Rational decision-making is based on true ROI considerations.

He added, "In nine out of ten times, ROI is a consideration in making a financial decision." but not the primary consideration. In technology acquisition, ROI may not be emphasized at all. ROI is often a secondary consideration. According to the Standish Group, 70% of IT and e-commerce projects either fail or are completed over-budget with less functionality than planned. Poor or inadequate consideration of ROI requirements may be a primary factor.

"ROI calculations should only be developed through an application of a scientific method. In order for an ROI to have meaning, you must know what the starting point is and to what standard you are trying to measure," said Greenbaum. He continued, "Most companies don't have good data for calculating ROI. Few companies have taken a financial picture of their current financial situation and don't have any ideas about where they are going." A major problem with ROI is that often-appropriate measures are not taken upfront to establish a baseline. If no baseline has been established, then there is no way to track ROI correctly. The bottom line is that you need initial, ongoing and concluding measurements to measure against the initial start date to properly calculate a meaningful ROI.

In many cases an ROI is not valid. The data isn't there to support it, there are no metrics, and the ability of the corporation to use metrics, which have been determined after the project is underway or completed, is often useless, worthless or meaningless.

Always remember, garbage in = garbage out. Collecting and interpreting meaningless information will of course give you meaningless results. Of course given the pathology of some companies, as described above, you may very well have been asked to construct a science fiction based ROI to satisfy the whims and irrationality of some whacked out senior executive. If so, good luck to you--make two envelopes.

What Makes Good ROI?
According to Greenbaum, "Often the major reasons that projects are undertaken in the first place is on the basis of relationships: between employees and senior management, a vendor and a company, a new CIO with a mandate and a previously existing, established relationship with a contracting company, vendor or consultant. Frequently executives are stuck using software and technology they may not understand, acquired through a merger. In many instances technology is acquired almost on a whim where the implications, financial or otherwise are not considered until after the fact. 'Hunches' that the product is great, can't be substantiated through numbers. An ROI can't be completed based on a guess. This can have both positive and negative implications. For Chrysler's PT Cruiser, the data did not support building it, yet this car is phenomenally successful by any measure."

Another factor in ROI is time to market, says Greenbaum. He added," If the time to market is too long, the company may miss the entire opportunity window for entering the market. Six months to market may make the difference between a company being the market leader or the market chump."

A metric-driven management style may also be both a major asset and a major liability at the same time. There may be something of a Catch-22 implicit in a request for an ROI determination. On the one hand, a company may need data to justify decision-making. At the same time a company may not want to use too much data because they may be afraid that they will be called out on the carpet for the data provided. A senior manager may also require you, the project manager, to build an ROI model as a CYA for him or her (How many envelopes do you need to make?).

Greenbaum's ROI Rules of Thumb
If you don't at least consider the ROI considerations detailed below, then you can't really justify what you are doing. Consider the following:

1. Calculate what is the actual requirement: Is this a business decision or a technical decision. It is the job of the executive to identify the need for the specific technology or business activity to be considered.
2. What is the problem today?
3. What is the cost to solve the problem?
4. What does it cost to build or offer the service in house?
5. What does it cost to build it or to offer it outside the company?
6. Are their technical issues to be addressed?
7. Can direct sales or other income be attributed to the purchased item and its supporting structures?
8. Will use of the item result in direct savings to the company or from improving related company processes?
9. The following costs must be identified:
10. 1. Cost to acquire, maintain and upgrade
    2. Cost to use
    3. Cost to train
    4. Can the expense be justified on the basis of cost or am I or my boss pals with the sales guy and we will buy it no matter what?
    5. Do I like to purchase technology or pad revenue for no particular business objective? If so, you can't make a true ROI.

What to Do if You are About to Open the Second Envelope
Let me see if I can paraphrase the following colloquial expression so it can be printed in a family-oriented website such as gantthead.com. Here we go. If you have to clean up poop, it's hard to come out with no poop on you. Was that clear yet relatively tasteful? In other words, what do you do if you are screwed by your company and left holding the bag in assuming a project midstream and have to manufacture an ROI after the fact, or placed in a situation where you don't have the information to build a true ROI model but are still required to build one? Either of these situations or many others with an unclear path to effectively determine an ROI can result in a lose-lose, career-limiting experience for some project managers.

Whenever you can lean empirically on the use of "meaningful" data, the more you can do and the better off you are. You may have to (and can) do the impossible. Try, if you can, to get ahold of people who made the original decisions on the ROI model. Short of giving up the ghost, try to find a way to perform some kind of a bold move to determine a previously hidden cost justification in the project. At the very least, you must determine the best and worst case scenario and calculate from there.

Oh yes, it is of critical importance, and I emphasize the word CRITICAL to clearly state all your assumptions and the derivations of any calculations in your ROI model. This way you can clearly show the origins of all of your calculations and defend them if necessary. Also see if you can get the (psychotic) powers that be to buy into or accept as many of your assumptions as possible. You want to lay a path for acceptance for your final ROI calculation regardless of whether it makes any sense at all. If it is a huge work of creative fiction, but your higher-ups have bought into it, you will then have your crack-addicted bosses to blame.

Unfortunately, these kinds of business practices are the way of the world. There is no such thing as perfect information or perfect data. To save yourself, you have to realize that virtually every decision made on the planet is made working from an information deficit situation, where people are doing the best they can with limited, imperfect knowledge. Always, as a professional, it goes without saying, do your best. You have to be the best possible employee and do everything possible to turn lemons into lemonade. You have got to find a way to turn a sow's ear into a silk purse. Those are all the cliches I have right now. Good luck and G-d speed.

P.S. If all else fails, make two envelopes.

reviews

"Nothing like calculating an exact science and latter qualifying that with going back and checking how accurate last assements have been. I rather enjoyed the article and its two envelope hook. I did an ROI Calculation that I felt quite proud of, an audit indicated that I was within $0.025 of the unit cost I calculated, fortunatly at 200,000 units I was low on my calculated cost and didn't have to open the second envelope. This wasn't entirely the case but it does go to the point of establishing a base and choosing the parameters to be used when the audit is conducted so that they are consistent with the origanl goals and objectives. I my case the per unit cost actual dropped $1.24/unit so a 2-3 cent difference either way had less of a sensitivity impact on my decision. Fortunatly, it came out the way it did and I looked to be an even bigger hero. I would say that if you drop poop on the road don't be surprised if on your return you find flies. "
David Rutchik - Jul 24, 2003