ESCAPE VELOCITY

1 - I think it's like the price of a diamond...the relationship of price to size as the stone gets bigger is not linear, it is more exponential. Either way, the cost per customer employee definitely varies by segment...but it varies in part because of vendor coverage models. A large vendor doesn't have salespeople to cover every 20 user opportunity, and a small vendor is less likely to be scaling up to 50,000 (though it can happen, either way, of course). I have 46,000 active Notes customers -- obviously I can't visit even 1% of them face to face in a given year. There are thousands of IBMers, though, that provide that economy of scale by representing Lotus.

Any which way, there are lots more opportunities at the small end, which is where IBM (like Microsoft and Oracle and SAP) are "channel-driven", or partner-led. The data makes the opportunity there quite clear.

2 - In a way, this is similar to the fundraising and campaigning efforts of a political candidate. Traditionally a candidate will be drawn to "special interest" groups because it costs less (time, energy, capital) to convince one representative of that group, and suddenly they've secured the vote and/or donations of each member. Going door to door and talking to each constituent individually just isn't as efficient. But to some extent they're then a slave to those groups because a change of heart from a single person can potentially finish them off.

So, from a sales perspective, if it's a choice between wooing the CIO of one Fortune 1000 into purchasing licenses for 80,000 employees... and individually pursuing 16,000 5-employee business owners, the natural choice would be to just woo that one CIO. The revenue would be the same, but the profit differential would be laughable.

On the other hand...... once you've got 16,000 customers with 5 employees each, if one of them drops you, you're out $500. No big deal. But if your typical customer has 80,000 employees and just one customer wanders off, you're out 8 million. Ouch.

So I can see some wisdom in making a gradual transition toward a focus on SMB: maintain the "easy" revenue stream, but work toward a less volatile situation, where the ebb and flow of a few customers here and there doesn't drastically change the landscape.

3 - @2 - A nice summary. The one thing missing is the fact that one, two, or who knows how many of those 5 user companies will grow. I have one company I'm working with now that had 200 users 5 years ago. They're at 1400 now, and looking to grow at least as much in the next 5 years.
The growth potential is something to consider in your equations.

4 - When I saw Charles' link I thought about the same analysis - glad you had a chance to look at it before I did.

Another way to look at this, though, is that the cost of sale is much more of a fixed cost than a variable cost, although it would be tiered (as a 100k seat sale will be a lot more involved than a 500 seat sale... but a 500 seat sale wouldn't necessarily be that much different than a 200 seat or a 1000 seat sale).

Looking at it that way and on a breakeven basis, if you take as an assumption that $100/user revenue figure, you get a breakeven on the cost side for the 20-1999 segment at around $11,000. Relying on a direct sales force (whether that resource is owned by IBM or by a partner), you're going to burn through that $11k pretty quickly. To make those sales profitable you need to rely on a lot of inside/telesales, customer pull (vs. push) and self-service information, without a lot of hand holding or as much direct involvement from a sales standpoint.

5 - First of all ... EWWWW with the tables and numbers on a Monday. Please, make it stop.

I think the idea of making money in the VSB space is really hard for a BP. Because of the size, you have to sell to more folks to make the same amount of money as to fewer mid and large sized companies. That means more marketing, sales, support, etc. The cost vs profit ratio gets warpped. I am not saying it can not be done, but its not a slam dunk market. And in 2007, the VSB customers expect the service and attention of a company much larger than them.

I know the projects that we have done in the VSB base is specialized, high end requirements. Things that make the company totally different than their competitors. Set them up for growth. We do very little with what I call the 'lifestyle VSB' - they are under 10 people and never plan on growing.

You need vision and a willingness to go outside the lines with the VSB space. Or you need a retail low-cost, low-support product.

6 - um... that should have been - 199 users in #4.

7 - @1 - Ed, I think one of the limiting factors for IBM here goes toward how that VSB market gets addressed. If the expectation is that the vendor has to go to the customer, then the cost of sale looks huge. But if you can shift the process to one where the customer comes to the vendor (which is really a retail process) then the cost of sale shift dramatically.

That, I think, is why you hear a lot of BP's complain about IBM's "marketing" efforts. Their desire is to create an atmosphere where prospects are coming to them, rather than vice versa. Because in a lot of ways, that's the only way for that segment to pay off big. (And they really mean advertising, which is obviously different.)

Microsoft really does have a substantial advantage in that space precisely because they are seen as a retail vendor for VSBs. (Yes, that has a lot to do with already being present on the desktop.)

IBM and its partners can only effectively pursue that market when the cost of sale comes down DRAMATICALLY, but the potential of that market could literally be a DOUBLING of software revenue.

That's why I express concerns over things like the transaction costs of doing business with IBM. If it's expensive to do business with you, then the customers won't come to you, and you have to go chase them down. And that's how your sales costs end up being enterprise-scale.

All that being said, I don't have good specific remedies, except to say that the more attention and effort, the better. And that IBM should remember that the value of strong branding is to get customers to chase you, rather than vice versa.

8 - @7 - totally agree. Thats exactly the difference between selling to a small and a large customer.

However before we go bashing IBM again for not paying enough attention to SMBs, i would say that it is a little unfair, because things have quite improved in this sector in the last year.

The whole express offering had an huge impact (we ONLY sell "express"). The new bundled offerings (Notes + Sametime) is a good idea and it is very likely that we will see a bundle with "Quickr" in the future. IBM even offers a nice bundle of 20-Notes Collaboration licenses + Server and client on site installation + Administrator training as a bundle for cca. 6000 USD.{ Link }
This is all targeted ONLY at (very small) SMBs !
And a major hardware distributor here adds to all eSeries-servers a license and installation pack for Domino.

Three years ago we had none of this !

So yes i too would love to see ...
- more advertising
- free versions for students/schools
- free Domino Designer
- images (VM-Ware) with preinstalled server applications,
- easier installation setups,
- lesser hardware requirements,
- lowered prices

...but on the other hand, as mentioned above, I am already quite pleased with everything i got in the last year.
Way to go IBM !!! (But please don't stop )

9 - Great analysis, Nathan. I seem to recall your background is in statistics or economics so I'm not at all surprised you were able to crank this out so quickly.

It's obvious that IBM has recently decided to start devoting more resources to the SMB space. The packaging and delivery have me concerned, but at least they're moving in a direction that shows someone understands they need to change what they were doing. It's also obvious that the potential in the 20 - 199 user space is huge but the cost of getting there is relatively high.

To your point about customer self-fulfillment, this is where better websites from IBM could play a major part in drawing potential customers in and lowering the cost of sale. In another thread on Ed's blog { Link } someone asked about the system requirements for the new Starter Packs, and Ed referred the reader to the regular Domino and Sametime requirements. It's one small issue but it's the little things like this, repeated hundreds of times, that make the web sites of IBM and Lotus a chore to deal with.

10 - Not to take it off track or be too defensive, but since both products' data sheets are available in one click from the product page, I am not sure it needs yet another data sheet of its own. But maybe I'm thinking too much like a large enterprise :-P

{ Link }

11 - @9 - Thanks Charles. And yeah, it's econ.

What I can't figure out yet is how a retail model for that space can dovetail into a BP program. I really have no idea if/how other companies do this. Microsoft is obviously big in the 20-199 space, but do they do that through partners? Or is it just through a retail perception where customers come to them? Do MS partners really only make money on customers > 200 people? I have no idea. I know that it's generally not profitable for even a small partner to chase business in the VSB realm, unless you can sell them on some big ticket items that really define their business.

Let's note that my input numbers are SWAGs here. I really have no idea what the revenue potential is of the 20-199 employee market, but at $100/head, it looks like a $6.5 billion market. That's obviously attractive. But is a software company ultimately completing with salesforce.com and gmail in that space? If so, it might be that the only way to get there is massive retail infrastructure investment, on the scale of those guys. Not selling software (or "solutions") but selling a service.

Is there a retail market where we can compare the process? Is it like buying a car? (Is a dealership like a business partner or VAR?) Is it like shopping for clothes? (Do you get your personal style representative when you come in the shop?) Is there some comparable model with the size/scope available? Or is it simply inevitable that attempting to serve the VSB and International Corporate audience at the same time simply leads to Microsoft-like conclusions, where you ultimately implode under your own company schizophrenia?

I dunno. The only thing I can tell is that the 20-199 market is brand-aware almost exactly like the consumer market, so a lot of the strategy has to be the same.

12 - Unfortunately, most BP are unprepared to venture into the VSMB space. Yes, the budgets of these companies are much smaller than the enterprise businesses. To be succesful in this market, you need to be offering services and products on a volume basis and that you can replicate over and over. This help reduce the cost of sales. What a VSMB spends on IT services vary greatly depending on the industry they are in.

13 - @10 - Ed, the data sheets don't have the system requirements.

@11 - I was thinking the same thing. To crack the VSB market is going to take a tremendous amount of advertising and money to build mindshare. MS has that because pretty much everyone who has ever used a computer has used Windows, and even the people who use Mac's still know who MS is. Everyone knows who IBM is but the Lotus brand is dwindling in recognition outside the IT community. Rebuilding that is going to be a tough uphill slog.

Even though they're SWAG's, it's still interesting to see it laid out like that. I have been looking at this data for a while and trying to discern something meaningful, and while I had some gut feelings I'm mathematically challenged and never got very far with an analysis of any type.

14 - @12 - Richard, I don't understand. How will servicing on a volume basis help reduce the cost of sales in the VSB market? It's precisely volume that has a high CoS, because the only way to get volume is to have LOTS OF CUSTOMERS, and each firm has a baseline cost of sale regardless of the volume of purchases they make.

That's my point about needing a "retail model" to work with that segment. The customer has to come to YOU to do business, and your sales cycle has to be at the worst no more work than it takes to sell a car. Heck, if you're going to charge retail type rates into a company of 50, you might need to have automotive-type financing options!

Actually, the more I think about it, the more intrigued I am by that concept. Could you have a retail style IT "dealership", with an inventory "lot" and sales specialists who help you get the most out of it? Service plans, warranties, financing, options packages -- that's an interesting mental picture, at least.

If it was like automotive dealerships, you'd have retail/wholesale pricing, vendor relationships with area-exclusivities, dealer certifications, all kinds of arrangements.

I wonder why no one's every tried that before? I bet it would make it incredibly easy to target that almost-consumer segment with a real local presence.

15 - Some things you need to figure in...

Heavy discounts that go to the upper tier customers. Discounts pull down revenue per unit sold, but the largest discounts go to customers who are buying more than one product, so the overall impact on revenue/employee is hard to figure.

Support costs for the lower tier customers, as John already mentioned. Maybe you're already figuring them in cost of sales, but my guess is not enough.

16 - While Nathan's calculations might be perfectly valid for the direct IBM sales channel, I would like to point out that IBM has both direct and an indirect marketing channels. The direct channel is focussed on larger customers and the indirect channel is primarily focussed on smaller customers, so the cost of sale and ROI calculations for different market segments will be different (for IBM). Business partners also have a cost of sale, but because they have much lower indirect costs than IBM, their total cost of sale per customer is significantly lower than for IBM.

I feel that in order for IBM to successfully address the SMB market, they needs to revise their Marketing not just their Sales strategy.

This means a greater focus on simpler products (like Domino), better user interfaces, lower entry level pricing, more cost competitive packaging of products (they are making significant progress in this area), much, much more advertising, increased brand awareness and significantly more promotional activities (particularly for the press and consulting firms), much better support of business partners, more education initiatives for partners, more strategic partnerships with partners (and not just the largest partners or US based partners but also for smaller partners and Non-US based partners as well).

To jump to the sales strategy briefly, I feel that in the past, software sales have often been used as a tool to generate hardware sales by IBM sales people, and far too often IBM sales staff and middle management have been generally poor partners with their indirect channel partners, but IBM needs to break free of this cultural artifact if they are to rapidly expand into the SMB market.

Another area that IBM needs to improve in is software development tools that are made available to business partners. The SMB market needs more business solutions that work out of the box. Some of these solutions need to be developed by IBM/Lotus and other solutions need to be developed by business partners. All of these business solutions need to be much easier to install, more reliable and much easier to support. Because, unlike the larger end of the market, the SMB market does not have access to an army of technical experts. Perhaps IBM/Lotus can facilitate the development of these solutions by providing more templates and more solution frameworks that can be "filled-in" or adapted to other market niches by business partners. And IBM/Lotus can also extend their "thought leadership" with additional open source software initiatives targeted specifically at the SMB market segment. Of cause, simpler, more reliable and better software will also benefit larger customers, so everyone will win from greater focus in this area.

But as previously stated, I feel that the biggest challenge that IBM/Lotus faces in the SMB market is brand awareness. Many people, particularly the younger generation are simply unaware that IBM/Lotus exists. And many that are aware of their existence see them as an obsolete dinosaur from a past era. That is why I say that the biggest issue is marketing and not just sales. But this problem can be easily fixed by IBM/Lotus with something that they have, lots of money.

17 - So I went back and looked at the US SBA numbers here, and Palmisano's statement doesn't make any sense.

Employees in enterprise companies (10000+): 30 million
Employees in large companies (1000-9999): 20 million
Employees in IBM-defined SMBs (250-999): < 14 million

So, how would it be IBM's largest market? By getting spending/employee to be DOUBLE what it is in the enterprise market?

Unless Sam's plan is to concede massive amounts of the enterprise market, his statement is pretty much impossible. Given the difference in IBM's penetration in those top tier customers vs. the SMB market, I just don't see how this pronouncement makes any sense. Maybe, if IBM were genius at this, they could see the single largest REVENUE stream come from there, but it would be far from the most profitable group, as the cost of sales would be through the roof.

Cross-posted back at Brill's

18 - Ah, I get it...

"which Palmisano explained at a compound rate would allow the SMB segment to surpass the financial services sector as IBM's largest customer."

So SMBs are all lumped into one sector for this defition, whereas larger organizations are divided along enterprise lines.

Because I guess as soon as you cross 1000 employees, your needs become industry-specific, where at 300 employees, every organization is the same.

19 - @14,

I am confused. the retail market is primarily a volume sales market unless you are at the high end part of a market segment. In order to achieve the sales volume, you need to use an indirect sales model so that your cost of sales is lower whether it is for services or products. Even if you are using a indirect sales model, if you can not provide services and products that are more turnkey your cost of sales and support will kill you regardless of the size of the deal.

@16
I agree with you. Sadly, I just did a survey with a number of small IT consultants that I know and asked them about the brand awareness of IBM and Lotus in the VSMB. Their answer was very little and in some case none. Ed was surprise when I mentioned that there were small businesses that have never heard of IBM a few blogs back. IBM should not be surprise since they do not advertise or create brand awareness in small businesses market. What they do have is too fragmentated and unfocus.

With the exception of Lotus Organizer and Lotus SmartSuite, most IBM and Lotus software require too much support to get up and running for a small business. In my opinion, even Lotus Notes and Domino do not meet the requirements. Just as IBM has begun with the user interface, IBM needs to review the installation processes and simplify and simplify. I recently asked IBM this question and I did not get a response. With the new messaging and collaboration starter pack, is the installation all "integrated" on a single installation CD? If it is not the case, then it is not ready for primary time for small businesses. IBM needs a whole culture change even if they are marketing to what they now call SMB. It would not surprise me in a few years that the IBM CEO announces that VSMB is their next target market for growth.

20 - @19 - It's a volume sales market in the sense that you need to have a high volume of sales in order to be successful. But you won't make many VOLUME SALES. Make sense?

Using an indirect sales model simply shifts the burden of the cost of sale, not eliminate it. You're simply pushing the cost on to the reseller, who will only be interested if they capture a larger share of the revenue, hence wholesale/retail markup.

The software business partner model is predicated on the idea that the BP captures a complementary revenue stream from the original source software sales. If I, as a BP, sell 1000 licenses of Notes, I'm not going to make squat off my share of that license sale. What I *am* presumably going to do is sell services with those 1000 licenses, and those services will be where the money comes from.

The problem is: this doesn't work in VSBs except in rare cases (like John & College Bowl), where the software/sales combination isn't just an enabler of the business, but actually DEFINES it. But if I build a great CRM package, when someone buys it, the package isn't their actual business. They purchase it as an enabler to do their business. And therefore they aren't going to want to invest in services coupled with it. Which means there's no incentive for an indirect sales model. Which means I can't defer part of my cost of sale. Which means I have to find another solution.

That's the trouble with the BP model and accessing that 20-199 range. The software that best services that market is turnkey stuff. Either hosted (Google, salesforce.com), or instantly installable and deployable (MS products pre-installed on desktops.) We all ask IBM to make their products easier to install, but the closer they get to enabling turnkey operation, the less opportunity there is for the BP to get anything out of the transaction.

There is an ISV model, where additional layers of BP software are coupled with the overall package sale. But having done this 3 times in the Notes world, I can tell you that the eternal desire in that case is to simply make the underlying software invisible and sell your total solution without talking about the platform at all. In which case, all the brand awareness in the world from IBM isn't going to help YOU.

21 - Great work, Nathan.

But the most interesting comment is in your comment @7:
"Microsoft really does have a substantial advantage in that space precisely because they are seen as a retail vendor for VSBs. (Yes, that has a lot to do with already being present on the desktop.)"

I think it's worthwhile for IBM to chase the VSB segment, even if they have to accept higher costs. The problem is that otherwise those VSBs go for Microsoft's SBS offering, and then grow into SMEs, and that company is lost forever...

Still, this is very enlightening as to how these kinds of sales scale up/down. When combined with Joel Spolsky's observations on software pricing ({ Link } it shows how difficult it can be to market to the SMB segment without making the other segments of the market feel like they're being ripped off - despite the fact that IBM would still be making a loss in the SMB market even with pricing and licensing differences...

Thanks!

22 - Funnily enough, I remember having a conversation at LCTY in Costa Mesa about how I believed IBM should have a retail presence for Notes/Sametime. If you displace Outlook in the personal/VSB space then the "I want Outlook" argument in corporate world can be weakened. You don't *have* fo use a server for Notes so you could go after this market by providing a single user configuration useable in the home.
The way the IBM Lotus software is currently sold is not conducive to this becoming a reality.
I reckon the only way to generate market share in the VSB space is to have products on shelves in retailers - it would reduce the cost of sale, and can still allow VARs to provide services as there would be more companies using the product(s). The only possible downside is the licensing protection which would have to be built into the product.

23 - Notes always could be a replacement for Outlook, it can connect to an SMTP server. Now with the Productivity Editors, It can be a replacement for Office. At quite a discount. Now seems like the best time to push for a consumer version of Notes. Domino servers are not *required*, but there are public ones out there, and there could be many more. It wouldn't be worth a whole lot of investment, but it wouldn't take much either. Just make it available (put it out there as a possibility somewhere on ibm.com) and see what happens. If Sametime could do a Trillian kind of thing that would be killer to include it too. Even one single person coming up to Exchange and saying "but that's not like Notes!" would be well worth it

24 - Some general comments -- both retail and OEM (PC manufacturer distribution) have been suggested. Have to say both are very very expensive CoS -- IBM isn't actively making any software for retail, so the "manufacturing" and "channel" processes aren't optimized for it... and OEM, well, OEM is difficult because every software vendor wants to be on the new PC, while only a few are really able to afford that approach.

Freely distributed products are a good idea in terms of consumer mindshare -- look at Firefox or Adobe Acrobat or whatever -- but I don't think they need to be boxes in a store or preloads.

25 - "IBM isn't actively making any software for retail..."

And just to be clear, I'm not sure whether it should start. The only thing that makes retail helpful is the immediacy of the perception of VSBs. Google and salesforce.com are good examples of ways to create that impression without brick & mortar retail.

26 - Yeah I'm not suggesting getting it on every shelf right beside MS Office (unless it proves wildly successful...), just make it available on IBM.com, a stand alone home office version of Notes or something along that line. Click to order, get a shrink wrapped box with manuals in the mail. Don't take out ads in every major newspaper and during the Superbowl or anything, just mention it in a press release and a few interviews, and see if there are any takers. Put any profit into building it up. This new Notes 8 Standard release seems like the perfect situation to finally give it a shot. Then people (IBM and/or others) could set up things like Google Aps or whatnot and offer them online *natively through the Notes client*. That might be fun It comes with enough built-in and with the included templates for local dbs to be compelling anyway, IMHO, but infinitely extendable by ASPs and whatnot as well. Just a thought, I know it's been brought up enough in the past, but it just seems like the game has changed now with the productivity editors, so it might be worth reconsidering:

Lotus Notes 8 Standard
-email&pim, word processing, spreadsheets, databases
price: <$150

Microsoft Office 2007 Standard
-email&pim, word processing, spreadsheets, databases
price: ~$400

hmmm...

27 - I am sorry to say but you chaps can see no further than the end of your nose. You are just looking at the US market as the only market AND the only people who can service that market.

I have a Notes business in Thailand (for over 14 years) and I would love to have lots of customer with a 1,000 seats plus. The fact of the matter is that the majority of customers here are in the 100 - 500 bracket. Of course our cost of sales is designed to meet that. However we often sell 150 - 250 seats to to customers and as long as we sell our applications and service we do very nicely thank you.

Companies like ours operating out of lower cost countries can easily service the VSMB in terms of sale of the licences and also product. Of course it would all need to be done over the Internet, which is not really a problem. This biggest issues to this is IBM its self. It makes it almost impossible for you to sell out side your own country, IBM needs to allow companies that are actually set-up to deal with VSMB companies to actually sell to them anywhere. We understand what they want and know how to deal with them. We have several customers out side Thailand where we even run there Notes/Domino systems remotely, via the Internet and get paid for it.

So come on IBM make it easy for SMB oriented BP to service your VSMB customer anywhere in the world.