Friday, February 07, 2025

Learning to Quote Profitable Jobs - and Three Mistakes to Avoid

Today's topic:   Learning to Quote Profitable Jobs - Lessons Learned


When I first started, I had lots of experience being the buyer of technologies. Of course, I bought lots of supplies, and the occasional small equipment. But I had also designed entire new office setups in remote offices and bought everything technology-related that went into the office, from telephones and servers to desktops and surge protectors.


Selling
all that stuff was another story. As the buyer, I was sometimes given a budget. But most of the time, I had to develop the budget, which meant that I had to fish around and figure out what needed to be included, then get some idea of pricing, and THEN develop a realistic budget.

As a seller, I normally designed the whole operation, and then had to figure out where to get everything. Then I had to figure out how to get a decent price. And, most importantly, I had to find the balance that would both sell the job and be profitable.

If I could go back and start over, this learning process would be a lot faster and a lot more profitable. 

Here are the three biggest mistakes I made:

One: The quote must have an expiration date. 

While sales people think this increases urgency (I don't believe it does), I learned that people will come back to you six or eight months later and want the old pricing. That's great if something dropped in price, but bad if it increased. Most often, however, the old equipment was no longer sold because it was based on last year's processor or memory.

Two: Volatile pricing needs an asterisk.

Around the turn of the century, memory chips fluctuated like a volatile stock market. Prices generally went up, but often went down. And something could double in price in a day or two. I remember a very big job that went from profitable to unprofitable because of the timing of when I had to buy chips.

You've been to restaurants where the price of some seasonal food depends on availability. You need to do the same thing if you can't guarantee the price of what you're acquiring. Just put an asterisk. "*Price not guaranteed until date of payment. To lock in this, price, prepayment is required." THAT creates urgency.

Three: Ignore what everyone else is charging. 

This is one of the most important lessons I've learned over the years. It applies to labor, contracts, managed service plans, hardware, software, projects, and anything else you can think of.

I mentioned this in an earlier episode, but we started out trying to charge no more than what the client would pay at a big box store. This was a futile and frustrating experience. As a new reseller, unknown to the distributor, and with no volume to speak of, my wholesale price was often higher than what I would pay at retail.

We finally settled on a simple formula:

Wholesale cost x 1.25 (for example, $100 x 1.25 = $125)

Thus, my profit is twenty percent of the final price (for example, $125 x .2 = $25)

Note: Some people just double their cost. So they pay $100 for an item and sell for $200. Good for them. That seemed a bit extreme to me, but it reinforced the lesson that you can totally ignore what anyone else is doing.

I set up my system so I get a decent profit and guarantee that I never lose money.

Some IT consultants argue that they can't sell for more than retail, suggested retail, or what the client would pay at Staples. That is a false belief and a self-imposed limitation! Ignore this belief! Tell yourself and your client these two truths:

1) We will only sell the right thing. We will do the research, find the right products and services, and create a combination that maximizes performance and lifetime value.

2) We warranty everything for at least a year, and our preference is to only sell business class equipment that's warrantied for at least three years.

In other words, you might be more expensive, but they definitely get value from it. If you sent the client to get something, they'd go to Costco and buy the wrong thing - guaranteed.  I'll never forget when I quoted a new desktop setup for a client and he said no. He went out and bought something that looked the same to him. It was the slowest new machine I'd ever seen. He said, "This is the same machine you were going to sell me." 

I said, no. The machine he bought had ZERO level two cache. I didn't even know that was possible. But it was half the price of what I quoted. It was junk. He was too arrogant to learn a lesson, but I got a story I can tell to demonstrate this point.

Bottom Line: If you sell the wrong thing, you might lose money up front. If you price it wrong, you will lose money up front. And if you sell the wrong thing, you will gobble up your profit down the road as you have to fix and repair something again and again.

Luckily, losing money is painful enough that I figured it out with just a few bad experiences. The lessons here are simple, but extremely important.

Bonus Tip: In your financial tool (QuickBooks, Xero, etc.) you should have categories/accounts for hardware, software, and services AND these should align with COGS - Cost of Goods Sold - categories/accounts for hardware, software, and services. Then, no matter what adjustments or weirdness happens, you have a clear set of numbers to compare. For example, Hardware sales should be about 20% higher than Hardware COGS.

What lessons have you learned when it comes to creating profitable quotes?

-----

This is Episode 17 of the ongoing Lessons Learned series. For all the information, and an index of Lessons Learned episodes, go to the Lessons Learned Page

Leave comments and questions below. And join me next week, right here.

Subscribe to the blog so you don't miss a thing.

:-)


Monday, February 03, 2025

The Two Biggest Challenges for MSPs - Solved!

I'm completing a course I teach on Customer Service for IT Professionals, and refreshing one of my most important slides - Facing your biggest challenges. In early 2025, your largest challenges remain unchanged. Luckily, the solution to both is the same. First the challenges.


#1 challenge for the next ten years
: Losing market share to well-funded large-and-growing regional and national MSPs. Whether you like it or not, the era of large MSPs is growing. Many, but not all, are funded by private equity. All of them have the goal of growing as much as they can. By any means necessary.

#2 challenge for the next five years: Artificial Intelligence. As Dave Sobel pointed out in his Business of Tech podcast, we are entering an era where you can ask your favorite AI chat tool to solve simple technology problems (See https://www.businessof.tech/podcast/debunking-the-myth-are-msps-really-bad-at-sales-with-ryan-morris/, timestamp 22:00-24:00 and after). 

Both of these are continuing trends. No one should be surprised. Now let's talk about the reality of these challenges.

Larger MSPs provide worse service. Period.

One of my fundamental beliefs about service is that larger organizations struggle to maintain quality service as they grow. There are many reasons for this. First, most business are not well run, and larger businesses spend a great deal of time focusing on the wrong things. 

Second, layers of management come silos of knowledge and fiefdoms of control. Every additional top-down command separates technicians and other employees from taking price in their work. People always work within a system, so when they build a system focused on something other than high quality service, they end us with a system that puts service lower on the priority list - by default.

Most of these companies are obsessed with KPIs that do not promote quality service or quality workmanship. For example, a metric based on how fast you can close a service request can only have a negative effect on quality of service. It can - and does - result in alienated technicians who hate their jobs, are stressed out, and couldn't care less about actual client-facing service.

These larger companies also tend to be highly focused on ever-increasing profit. Profit is more important than clients. Profit is more important than employees. Profit is more important than service. Profit is more important than doing the job right the first time. The only possible result is lower client satisfaction, lower employees satisfaction, lower quality service, and more rework.

The never-ending search for higher margins also means that these companies charge less and less. Because they cannot compete on service quality, they compete on price. That's bad IF you try to compete with them. You don't have to make that choice. 

Lowering their price means lowering their profits per job, so they have to force their over-worked managers and employees to work harder and accomplish more with fewer resources. Nothing about that equation results in an enjoyable company to do work for or with. Quality continues going downhill and job satisfaction goes with it.

If you could design a company to compete with, it would look like a regional or national MSP. 

Artificial Intelligence is your best friend.

As Dave pointed out on The Business of Tech, we are already seeing ChatGPT and other AI tools solving basic problems. This will happen more and more. Again, this is 100% predictable.

Today, if you want to learn something totally new, you go to YouTube and watch a video. "How do I remove the rear speaker covers on a 2003 Toyota Camry?" Boom. Not one video. Lots of videos. And down the rabbit hole you go.

Very soon, you'll ask your questions of AI. "How do I serve DHCP on a third VLAN with an old HP switch?" A few follow-up questions and you're good to go.

The next step - coming in 2025 - is for dentists and accountants and their office managers to ask the same questions. This gets to the most important truth you need to remember about AI:

AI will not take your job. Someone using AI might take your job.

And the very, very good news is: You already know how to stay ahead of the game. You just have to be willing to do it. If you haven't read The Greatest Secret in the World by Og Mandino, today's a good day to start. (See https://amzn.to/3ExqxBB)

The greatest secret? It only takes a small, consistent effort to be better than everyone else. Why? Because most people will not make the effort. They will not be consistent. They will not do the only thing they have to do to be excellent and separate themselves from the pack. You can.

The solution to both problems is the same. This is very good news.

How do you win in an environment that is dominated by growing MSPs trying to steal your clients and artificial intelligence making tech support easy for everyone? It's a simple two-step process.

1. Be supremely competent

2. Deliver extraordinary service

Competence will always win the day. The greatest challenge medical doctors faced in the last fifty years is the democratization of knowledge. I can Google my symptoms and self diagnose a lot. Most of the arcane knowledge "owned" by physicians is now widely available on the Internet. 

I still have to figure out when and how to use it. But doctors are no longer considered god-like beings with special knowledge. No job is more threatened by AI that doctors. But when *I* need surgery, I don't look for the first available quack with a parchment on the wall. I want the best doctor I can find. I want someone who has "seen it all" and is recognized in their profession.

You must be the same way. Basic problem solving has always been the lowest level of IT consulting. When no one knew how to build a network, those who did thrived. When no one knew how to set up remote services, those who did survived. Slowly, our specialized knowledge became easier and more widely available.

This will never end. This is our industry. This is the life you've chosen.

Soon, much of the "Level One" support you offer will be available to nine-year-olds on their smart phones. Being really good at that will be essentially useless. You need to focus on all the really hard stuff that AI won't figure out for a long time - or maybe ever.

Those who've been in any industry for a while know the advantage of having seen a wide variety of problems and rare issues. At least for the foreseeable future, AI sucks at integrating all that. Be supremely competent and you will always have work.

Let other companies turn over their first tier support to chatbots that are already the alternative to your lowest-level challenges. Focus on things they can't do.

And deliver great service along the way. Remember: Service is Everything You Do. If your entire business is built around giving amazing service, you will find clients who want that and are willing to pay for it. You will also hire the right people, offer the right combination of products and services, execute perfectly, and build a business you want to work in.

Five years from now, almost no one will pay good money for level one tech support. But almost everyone will pay top dollar for supreme service and piece of mind. Your client never asked for technology. They don't care about it. It's not why they went into business. They do care about offering supreme service to their clients. Helping them do that will keep you very profitable and busy for a long time to come.

Supreme competence + supreme service = A Very Successful IT Business!

:-)


Sunday, February 02, 2025

Quality Service is Never a Mystery or a Mistake

One of the unfortunate results of cutting my technology teeth in academic and enterprise businesses is that I often approached technology from a purely technical perspective. For example, what color should the network cables be? Answer: Whatever fits the technical specs and is available in bulk at a reasonable price. 


For example, when a company like HP is wiring a building the size of a warehouse, no one cares what color the wires are.

... and then I started working in small business. I'll never forget the day that we were one office into setting up the network for a twelve-office organization. The owner came in and asked, "Do the wires have to be blue? Everything else in our office is beige, off-white, or earth tones. 

I grew up poor. We only had the box of eight crayons. My brain understand blue and red and green. So there was no chance that I would even consider looking for beige network cables. But I did know that we could get white. I offered up white cables instead.

The client told me that's fine. Then she added, "Don't you think white cables would look better at all of your clients than blue - unless they have a blue office?"

This fall deep into the territory labeled, I would never have thought about that. But it made perfect sense. As with so many checklists in our business, I simply added the question, "Do you have a preference for the color of the cables?" to job specifications.

“The essence of sustainable competitive advantage is:

1) The obvious;

2) The little things;

3) The accumulation of little things over the years.”

— Tom Peters

Too often, MSPs take the approach that technicians either are good, or are not good, with the details. When it comes to a new laptop setup, we give them a detailed checklist. Did you remove all the junk software? Did you apply all the patches? and so forth.

That list grows and gets fine-tuned over time. The goal is to make sure that every little thing is perfect, to the extent it can be. You never want to be driving away from a client's office and get an email asking why Adobe Acrobat wasn't installed. 

ALL technicians are good, if they work inside a system that's good. All technicians are haphazard when they work inside a system that's haphazard.

Most of the time, it's appropriate for us to focus on the technical. Do we have the right hardware, the right software, the right services for the job? Are things configured correctly? Did we do everything well, in order, and document it? Check. Check. Check.

But from the client's perspective, there are lots of non-technical considerations. Back in the day, selling big, clunky, ugly desktop computers was fine. It made people feel that they actually bought something worth the exorbitant price. Today, people want technology to be more elegant or even invisible. Part of your job is to make sure the job is esthetically pleasing as well as technically awesome.

How do you train tech-first employees to ask where a machine should go, how the wires should be managed, or the color of the cables? Give them a checklist! They are certainly capable of having those conversations. It's up to you to make sure they do have the conversation (or you do, or the sales person does).

As with everything else in your business, quality never just happens. You need to make it happen. And being consistently better at "the accumulation of little things over the years" is part of your brand. Are you the company that does sloppy work, always forgets one thing, and is plagued by re-work? Or are you the company that does awesome work the first time and makes the client glad they made the investment? Neither of those happens by accident.

Quality control is not a mystical art. Anyone can be good at it. You just need to make it part of your checklists from the beginning.

:-)