This is On-Choosing-to-Write-a-NeXT-Application in view mode; [Up]
Date: Sun 28-Sep-1991 03:07:03 From: thomsen@spf.trw.com (Mark Thomsen) Subject: On Choosing to Write a NeXT Application A thought exercise ... The decision to write an application for a machine is influenced by the number of potential purchases. Given that not all machine owners will buy your type of app and given that there might be other apps of the same type, the potential purchases is approximated by # machines x % with this type of app x % share of app type for your app M x A x S So a young company will look at the # of machines (e.g., PCs vs. Macs vs. Suns vs. NeXTs). They will estimate how many of those machines are potential sales for this type of app (e.g., databases). They will figure out their share of this market (machine + app type). The decision to write an application for a machine is also influenced by the investment required to develop the application, approximated by: # people x cost rate per person x time to market P x R x T The decision is also influenced by the investment beyond development, including advertisement, trade shows, and distribution. There are other costs, but let's just work with these for the moment. Now, the total # of machines is important of course, but consider - If M is large and A is low for one machine, and M is modest and A is quite large for another machine, then the second market might be equally attractive - same MxA. - If M and A are large (e.g., PC spreadsheets) then probably there are many alternatives on market and in development; it might be very hard and long to get your share, S, to be significant. - If MxAxS is twice as good for one machine vs. another, and PxRxT is twice as long for the first machine vs. the second, then the question is return on investment and other ratios. Also there is increased risk if the T is larger - more products can come to market and get upgraded. Also, with a large MxAxS the issue of marketing and distribution costs become critical - and again factor into return on investment. - If PxRxT and MxA are both large, and S is not huge at the outset, the cash flow will be very red for a long time relative to many alternatives. The investment profile has to consider this and might avoid that market, even if the long term MxAxS is superior to alternatives. So, if you can only get modest capital before you get to market (e.g., you invest your own money), developing for a NeXT might make sense. You might be first of kind. Porting apps without real value added is not going to capture the market share in the long run so you might get market share. Your time to market and total investment might be less (averaging various development efforts we know of, time to market is half on the NeXT what people see on PCs and Suns, and the investment is a third). If the combination of NeXT+your app kills other combinations and provides unique advantages, NeXT may help you (ala' Lotus Improv) and your marketing costs for initial sales might be quite low. Large companies that can make large investments and that have the marketing and distribution channels in place will tend to focus on large MxA. If you wait for assurance of large MxA (e.g., Microsoft's stated strategy) then you are reactionary relative to new markets. If you can get in and lock a large S, and if MxA starts small and grows, you can do very well. Some large companies will plan ahead, betting on the MxA growth and locking in the S. Small companies, if the PxRxT is low enough, will take this route as well, if the cash flow works out. For NeXTites, looking through various return on investment paths, one can conclude - many of the borderline and some larger companies will wait until the NeXT market grows (Borland? Microsoft? Quark?) - their investment profile requires large MxAxS, and they are already investing in such markets - some companies will invest with an advanced product (Lotus w/ Improv) - they want to dominate app type (large S) on NeXT and hope MxA grows, - porting dominant products will occur infrequently (WordPerfect is an exception - we don't see 1-2-3, FileMaker, Word, et.al. coming to NeXT) - people not in large MxAxS markets will consider going to NeXT (e.g., Oracle, Ingres - their main market is Suns and up) - the return on investment is similar to how they already operate, and especially attractive if they can do a port without having to invest a lot to add value - most of the products will initially come from small companies - they need some return on the investment they can make, and they will be impatient (for realistic reasons) to get to market to get cash flow in the black Bottom line - yes, there is some software and companies that won't be coming to NeXT real soon, but the situation is quite rich and it is highly likely that if the MxA is at all interesting on NeXT then some company will produce that application and it often will not come from the expected sources. I wrote this after reading the Philippe Kahn interview in a recent Upside. His 'barbarian' environment intrigues me as a model for keeping growth from altering your decision making flexibility. If you bloat you are less able to justify smaller markets and riskier moves. I am impressed with the thinking. I also went back to Bill Gates talking about the only two costs a software company can really control - people and marketing. Mark R. Thomsen
Date: Sun 28-Sep-1991 04:07:12 From: mcgredo@bluto.ie.orst.edu (Don McGregor) Subject: Re: On Choosing to Write a NeXT Application Mark Thomsen writes >A thought exercise ... > >Large companies that can make large investments and that have the marketingand >distribution channels in place will tend to focus on large MxA.If you wait for >assurance of large MxA (e.g., Microsoft's stated strategy) then you are >reactionary relative to new markets. I basically agree with this. The real, innovative apps will come from companies using the NeXT as a sort of prototype machine; they do something spiff, and sell it. You don't need much marketing muscle to get the word out in the NeXT market right now, so the small players are on a level playing field with the big boys and you won't get lost in the shuffle. Low development costs mean the cost of failure is low. Once you've established yourself on the NeXT you can move to something else and work on that while you've got a regular paycheck coming in. In some ways, Quark is doing this on a grander scale: they did their first product on the Mac, and are porting it to the high volume Windows world. Microsoft had been doing it for years with Excel, Word, PowerPoint, and MS Mail (though they seem to have moved primary development back to the DOS/Winodws world again). It's also the explicit strategy of Pages corp.
Date: Sun 04-Oct-1991 23:27:57 From: thomsen@spf.trw.com (Mark Thomsen) Subject: Re: On Choosing to Write a NeXT Application Don McGregor writes > It's also the explicit strategy of Pages Corp. (It = small company getting > something new ("spiff") started on a new machine to get in early). Exactly true, as true for LightHouse and some others. I cannot imagine the b***s at Lotus to produce Improv - big company, phenomenal MxAxS with Lotus 1-2-3 on PCs, and they do something radical on a new machine. That is really impressive. Ashton Tate looked like they were headed in that direction but corporate management pulled them up short. Mark R. Thomsen
Date: Sun 12-Oct-1991 12:46:48 From: Unknown Subject: Re: On Choosing to Write a NeXT Application In article <28ECF8FE.4ED@deneva.sdd.trw.com> thomsen@spf.trw.com (Mark Thomsen) > Don McGregor writes > > > It's also the explicit strategy of Pages Corp. (It = small company getting > > something new ("spiff") started on a new machine to get in early). > > Exactly true, as true for LightHouse and some others. I cannot imagine the > b***s at Lotus to produce Improv - big company, phenomenal MxAxS with Lotus > 1-2-3 on PCs, and they do something radical on a new machine. That is really > impressive. Ashton Tate looked like they were headed in that direction but > corporate management pulled them up short. That may change now that Ashton-Tate has a new master, one with vision... Phillipe Kahn (actually, has the buyout actually been completed yet, I think Ashton-Tate approved it, not sure about the government)
These are the contents of the former NiCE NeXT User Group NeXTSTEP/OpenStep software archive, currently hosted by Marcel Waldvogel and Netfuture.ch.