But first, let’s define production. In any mid- or large-size project, you need to have different copies of your data, software, and maybe even hardware, which we call “environments.” All companies have a production environment – the live data where payables enter invoices, cashiers make sales and people enter expenses. Then, depending on the size of your organization, you may have other copies as well. Companies often have a development environment (for programmers to work in), a testing environment (for testing changes), and sometimes a training environment (which should be an exact copy of the production environment).

It’s important to have these environments. If a programmer is trying out new functionality, you don’t want him or her interrupting daily work. And while it’s less of an absolute rule, it’s still better to have programmers and users testing in different environments.

Because you have these different environments, you’ll never have to make changes in your production environment without testing and testing again elsewhere. Makes intuitive sense? But like many rules, it’s too often ignored in practice. Here are a few reasons why:

1)    It takes time to keep environments up to date. This happened at a billion dollar client of mine.   They had separate servers for production and testing for their ERP software, but because they hadn’t made significant changes in some time they didn’t bother to keep the testing environment up to date.

2)    It takes time to formally test. Sometimes companies think they can save time by making changes right in production. Wrong.

3)    It’s more expensive. This is often a sore point for smaller, growing companies. If you’re used to having one copy of QuickBooks and simply restoring it from a backup if it crashes (you do have a backup, right?), the idea of creating alternative environments may seem expensive and unnecessary. Large companies will usually spring for at least one alternative environment initially, but they tend to lock down their wallets when it comes to keeping them running or up to date. They’re also often reluctant to create new alternative environments even when it will improve efficiency (such as keeping developers and users separated).

4)    Management doesn’t understand the need. Again this happens often with rapidly growing companies. People who aren’t technical and who have an “I only have one set of QuickBooks” mentality, don’t know what they don’t know. I once had a client who ran hundreds of restaurants. They were always making changes to their register programming, and they didn’t have testing registers to confirm changes before deploying them to the field. Trouble.

We forget that the rule of never testing in production should apply to every piece of software you use. Are spreadsheets part of your budget process? Then you need to have locked down versions. I’ve just finished a budget process where 20+ units all had slightly different spreadsheets to combine in a “simple upload.” ARGH!

Is your company rigorous about never testing in production?

Leaving the meeting, I wondered how many of the finance executives in attendance are equally open within their own companies? Do they let their own employees know what’s going on? Or do they think (to quote Jack Nicholson), “You can’t handle the truth”? I’m guessing many of them don’t disclose as fully to employees as they do to the bank.

I often think people are only as transparent as they have to be. The bank can pull your debt, so you want to build a relationship. Generally, your employees don’t have as much leverage, but don’t you also want to build trust with them too? When the numbers aren’t clear, people assume the worst. I’ve seen this happen again and again with clients. When information is limited, people start to think managers are in it for themselves, even when there is no grand conspiracy.  With more transparent communication, there wouldn’t be as much grumbling – even when the news is bad.

I recognize this is something I need to work on myself. I have a natural tendency not to share information. I grew up in a small family business that drew a line between family and everybody else. I’ve been thinking about how to roll numbers out to people so they can take more ownership of individual projects. But I’m wondering how much to share. How much should your employees know about profitability?

Let’s start with how much writing I do. I post three times a week and my average post is 350 words. So, that’s about 4500 words a month that actually make it on the blog. For each month’s worth of posts, there are probably another five ideas I scope out but that don’t get posted. So, I’m up to 6000 words a month. Of the 14 posts or so that make the final cut, about a third I get right on the first shot. Another third, I redraft about half the article. The final third I wind up rewriting almost completely – keeping the idea but completely restructuring it. And this is all before the editor gets it. So, let’s say I’m writing 8000-8500 words a month for the blog. If you add in other writing I do for white papers and such, I’m writing about 9000 words a month on newsletter-type topics.

So, how many of them are really funny? Funny enough where people will come back to me months later and joke about it? I’d estimate that while we get a smile pretty often and, hopefully educate regularly, the really funny posts (judging by click-through rates) happen once every three months. So let’s say that’s 350 really funny words out of 27000 written – a little over 1%.

Which means, to my chagrin, I’m not a naturally funny writer. I just try a lot of stuff and sometimes it works. Which brings me back to my thoughts about focus. I can do good work without too much strain. But really remarkable work takes a lot of background effort. So, before I take on a new area, I have to ask myself: can I put enough work into it to get really remarkable results?

The hard part of making the jump from one set of books to multiple sets has nothing to do with software. You can maintain multiple sets of books whether you have a major ERP system or a smaller, midsized package. The hard part is accepting you won’t have “just one ledger” or “just one profit number.”

This has come to the fore for one of our clients. Because the client wants to keep to one ledger, we’ve had to book entries back and forth and twist reports to make things work. Our client is a U.S. company with satellite offices in the Far East. These foreign subsidiaries both generate their own revenue and do work for other parts of the company. Keeping to one ledger is difficult because they have two different audiences for their reports:

1)    Management: Management wants to attribute revenues and expenses to specific offices so they can make decisions about staffing and investment. They know some revenue is locally generated and some is just “intercompany” work they’re doing for other parts of the organization. They want their financials to show which offices are generating sales and which are just cost centers.

2)    Government: As anyone who has worked internationally knows, what’s logical for management has little to do with what’s logical for tax authorities. There’s an entire discipline built around transfer pricing and how much revenue to allocate to legal entities in different jurisdictions. Further, while operational management wants to divide profit a certain way, financial management wants to insure that excess capital isn’t tied up in foreign subsidiaries. For these reasons, statutory profit (or profit from the perspective of tax authorities) can be very different from management profit.

Software isn’t the problem here. The problem is accepting that there will be multiple profit calculations that will need to be reconciled. This is a major change for people who want every report tied directly to the same number. But given differing management and statutory drivers, there’s no choice. As Einstein (not Albert, his CPA cousin Herman) said, your books should be as simple as possible but no simpler.

Have you ever resisted a change because it makes things more complicated?

Before I go further, let me explain why this is a bad idea. All data and data tools are not the same. Broadly speaking, there are two kinds of data: structured and unstructured. Structured data is the stuff of accounting systems and the like. Every customer invoice has a customer, an amount due, and a date due. Every general ledger transaction has an account, an amount and posting data. This stuff is handled by databases like Oracle, SQL Server, and DB2.

But that’s not the only data out there. Word documents are data. Twitter posts are data. Pictures of flowers are data. But this data is unstructured – and, not surprisingly, it’s much harder (and sometimes less useful) to make it conform to a regular structure. Yes, Jane Austen and Charles Dickens are both 19^th century English novelists. But knowing that doesn’t really tell you much.

Lotus Notes was one of the first products to deal with unstructured data. It isn’t the kind of tool in which to write a general ledger. Indeed, you have to wonder why someone would want to write a general ledger at all when there are so many good software packages on the market.

Which brings us to the problem of “when all you have is a hammer, everything looks like a nail.”  Some techies fall so in love with a product or platform they want to use it to do everything, even when it isn’t suited to the purpose.

This isn’t a problem I come across too often. More often, people want to bring in too many technologies and then argue about distinctions that make little difference to their businesses. General ledgers can’t handle pictures of flowers, and one type of data software can’t handle all problems.

What strikes me is how often people turn matters of preference into matters of principle. Perhaps they want a boss who likes Excel presented a certain way. Or the company changes policies on something and employees say, “We haven’t done it that way before.” And I’m not talking about ethical issues or cheating. But it amazes me how people get caught up in the way things have been “always” been done. As in, “It was good enough that way for ten years, why are we changing now?” To which my answer generally is “There’s a new sheriff in town. You can follow her rules or you can choose not to. But she’s still the sheriff.”

As a consultant, preferences bother me less than most people. I worked with eight different clients in January. No two of these clients have the same standards or ways of communicating. I want to get paid, so I adapt to each client’s way of doing things.

But for many in-house people, having to adapt can be really hard. Their preferences become part of their personal sense of self, and when someone asks for change, they react negatively – in some cases even jeopardizing their jobs.

Why the strong reaction? Sometimes it’s a matter of prestige, sometimes it’s a matter of personal style; but it’s not a matter of principal. I can tell you, as the guy who talks to all sides, people’s understanding of what’s going on isn’t all that different. It’s just a matter of preference.

As I said, I’m not a business coach. But I can point out where preferences diverge from principles and sometimes that helps people move forward. But sometimes they need help from a real coach or therapist to fully accept the change.

Do you ever make a preference into a matter of principle? How did it turn out?

As I look at the situation, I realize we got in trouble because we didn’t follow a core rule: assign clear responsibility. We have no problem when our clients outsource complete tasks to us. And we have no problem providing advice and training and keeping our hands out of production (which we actually prefer). But in this case, we were neither here nor there. Administration was handled by a combination of client resources and our resources. Everyone understood part of the process but not all of it.

This had been going on for years without issue, so I didn’t push to fix it. But then during the last upgrade, testing wasn’t performed well. People threw their hands up saying, “That’s not my piece” and we tried to cover. Unfortunately, things didn’t go smoothly which created room for our misdirected email.

So, we’re stepping back and clarifying who has responsibility for what. Defining areas we own entirely and areas where we provide advice but don’t control the data. We’ll help, of course, but unless we have full responsibility, we won’t actually enter data (which we sometimes did before, just to get things done). This slows things down, but we’ve learned that short cuts just end up taking more time. And we’re done with that approach.

Have you found yourself part of a broken system? How did you deal with it? Did it get fixed?

I can think of three reasons for making your reports pretty:

1)    It’s part of making data visual. Even when we don’t have a catchy graph or a great picture to illustrate our point, people evaluate HOW it looks before they pay attention to WHAT it says. Once upon a time, computer output came one way – on large stacks of green bar paper. But today, everyone receives sophisticated output – bank statements, article downloads – and people expect it to look good. If your stuff doesn’t look good, it might be perceived as unimportant or (horrors!) boring.

2)    It makes things more clear. You know how programmers scoff at multiple colors? Go to a meeting and ask people to find column six. Then ask them to find the pink column. Guess which column people locate faster.

3)    It saves time. Has anyone ever sent you an Excel spreadsheet that doesn’t print nicely? If you’re skilled, you fix it. If not, you print it out and use a ruler. It’s aggravating. Nicely formatted spreadsheets save time and make you think more fondly of the person who sent it to you.

At Red Three, we’re still going to spend most of our time crunching numbers. But we should remember that getting the data to look good increases the odds of others actually taking the time to look at it.

1)    Ignore outside commentators who have no stake in your success. Here’s a quote I’ve used in presentations many times. It’s from November 2007, right before Manning took the New York Giants to victory over the New England Patriots in Super Bowl XLII (where Manning also received the Most Valuable Player award).

The Giants have to be worried about Manning again, both for the rest of this season and going forward. These [Manning’s] performances are inexcusable. Instead of asserting themselves as the NFC’s third best team and taking a big lead for the No. 1 wild-card spot, the Giants bounced themselves right back into the pack and are 1-2 in the second half of the season. 

NY Daily News – November 26th 2007 – Gary Meyers

I realize disparaging sportswriter prognostications is like shooting fish in a barrel. But how many managers and executives do you know who go around quoting from the book they’re reading this week – even if it contradicts last week’s book?

2)    Some numbers matter more than others. I’ve made this point before in seven ways to get people to look at your numbers. But it applies SO well during playoff season. As a friend of mine (and Redskins’ fan) posted on Facebook: “Wow. The Redskins beat the Giants TWICE this year! Which gets us . . . nothing!”

Bingo. If your goal is to win the big game, your regular season performance doesn’t mean much once you hit the playoffs. Yet how many people continue to focus on “regular season stats”? The same point applies to business. Figure out which numbers matter and tune out the background noise.

3)    Balance matters. The Green Bay Packers had a great record coming out of the regular season, but they continued to struggle with a defensive line ranked lasted in the NFL. When they met the Giants in a division-round playoff game, Green Bay quarterback Aaron Rodgers could no longer bail out their weak defense. The Giants took the game 37-20. If you want to be Super Bowl champs, you can’t do it with half a team.

The lesson for business? If you think your business is SO good in one area that it can make up for deficiencies in other areas, think again.

There’re two ways we’re helping our client show their work. First, we’re setting up separate accounts for the same liability for each currency involved. So, for example, instead of having one trade payables account, we’ll have one for trade payables USD, trade payables CAD, etc. The system doesn’t require this. It can perform the revaluation even if multiple currencies are posted to the same account. The problem is that the math becomes harder and it’s not easy to catch mistakes in rates and postings.

Second, (you guessed it), we’re building a custom report. The system’s revaluation report just shows one number and doesn’t show how it got there. With our customized report, we can show the beginning balances for an account in different currencies, the monthly postings in currencies, and the month end revaluation amount. One would think an expensive system would provide this out of the box. Sadly for my client, but happily for me, it doesn’t.