hi david
No need to get wound up with case studies - why not take a typical SME installation (I would suggest up to 4x peer to peer network with broadband, based around XP), and identify the incremental costs of on premises vs Saas for the basic financial ledger application. If you want to get sophisticated you could do this for different business types, although I suspect that would not add too much to the basic premise. I would suggest you split the costs between implementation and on going, and it would be fair to assume for on premises that there is an annual support contract in place (even though I am sure that does not apply to most line 50 installations!).
My point that TCO is not the prime purchasing driver is based on my experiences of SME's. Totally unscientific and unsubstantiated I would agree, but mostly they buy on recommendation - usually from the supplier of their operational systems. They are also reluctant to change -which is why the Sage acquisition model is (was?) so effective.
Pricing models etc.
David C,
Great.
On pricing it's fair to say that rental options have been available from some suppliers for a while, and larger enterprise outfits like SAP and Oracle are trying to move more of their customers to that style of billing (to make their lives more stable). The big difference is that in most circumstances there is a 1, 2, 3 or more year contract tie in, so it's a bit more like leasing than real "pay as you go" on short notice.
jc,
Your so right about loosely coupled services, and SaaS versus hosted with Citrix.
David Terrar
Mailto:dt@d2c.org.uk
web: http://www.d2c.org.uk and http://www.twinfield.com
blog: http://biztwozero.com
Hi David
"we need sceptics" .... Unfortunately you get sceptics whether you need them or not!
You can debate TCO until the cows come home - but it will get you nowhere. TCO is irrelevant to what you are selling. Dennis is quite right - it is the service bit that is important, along with the integration bit. You have your pricing structure, Sage has theirs - but its not price that causes the punters to bite.
Put it another way - you are offering a Saas product using the gartner definition - but you have a prospect who wants it on premises - would you never bite? Personally I would argue that the world is moving on, and traditional business models will start to fail, but you can't discount the open source community - and I bet that many will always want their own installations.
Response times
David,
That's a complete misconception. Decent SaaS providers have optimised their solution to give good performance on slower speed lines (assuming they've been around for a few years), and with current broadband you should have no issues whatsoever. We've got in excess of 22,000 subsribers on our system, and if you talk to any of them they'll tell you response time is not a problem, and exactly what they are used to from a desktop application on their own PC. If it wasn't, they'd soon move off to a better service.
We've got a SaaS partner called Furniture Software Limited who sell solutions to retail furniture shops. The key selling point to their application is a highly configurable and responsive customer order pad. Salesforce.com is the poster child for SaaS. There you have sales people interacting with customers as the core application. As well as SMEs working happily, they have enterprise customers with 1000s of users.
Lastly, I've demoed Twinfield to you myself at your place. If you remember we had no issues with response time during that session.
David Terrar
Mailto:dt@d2c.org.uk
web: http://www.d2c.org.uk and http://www.twinfield.com
blog: http://biztwozero.com
response times
no reason why reponse times should not be comparable to LAN, perhaps subject to contention issues, although that can equally apply to a LAN.
As David says, its down to how the software is constructed. I have seen browser based apps that are slow response, and equally some that are fast response.
The gartner
analysis is actually quite illuminating, and there still appears to be a tendancy to overegg the benefits compared with on premises.
I think Dennis is quite right to focus on the service rather than the software. And to that end there is not enough here on integration - under which model the user takes responsibility for assembling a collection of apps (however provided) to deliver a complete solution - somewhat different to the model this article articulates.
Its fine to talk about data portability, but how can that make sense if there are not open standards. For most users the cost of changing ledger apps is considerable and not to be undertaken lightly - so in practice what does a months notice really mean?
We need sceptics but...
Hi Alastair,
I'd be interested to know from what experience you believe anybody is over-egging the pudding? It should be relatively straightforward to do a like for like comparison, provided you take all of the factors in to account, although I have to admit that there isn't enough case study material with TCO or ROI data available. In our consulting environment we have 30 years of experience of ERP and finance implementations of all shapes and sizes. In the SaaS based implementations we've done so far, which have been supply chain collaboration, CRM and finance systems of varying sizes, we reckon the consulting and training involved has been around half what we would have expected with a conventional on-premise approach. Once you've added that to the economies of scale that the one-to-many model provides the end user, the comparison is very favourable indeed.
You are absolutely right to ask about integration. We were given a brief of 500 words, and John already let us push it to 900, so we couldn't cover everything. The better SaaS players realise that integration of various web services is vital, and so make sure it's easy to import/export data in various ways as well as providing online/real-time access to other systems through an open API. There will be examples of that in the use cases in the next article.
On switching - a month's notice means that compared to having paid for all of the hardware and software up front, thinking about changing starts to become economically viable, rather than a waste of 18, 24 or more months investment. Obviously a re-implementation cannot be taken lightly, but the way data can be exported from one service and imported to another makes the process a possibility to consider. Consequently, the SaaS vendors are much more focused on customer service day to day, compared to the vendors who got their rewards up front, and are on to selling the next one.
David Terrar
Mailto:dt@d2c.org.uk
web: http://www.d2c.org.uk and http://www.twinfield.com
blog: http://biztwozero.com
Standards/integration
Integration is an issue I've explored lightly at my place Alastair. Bog standard XML is OK to work with provided the vendor offers hooks or import/export.
I think you'll find that later this year, we'll start to see assembly style applicaitons becoming more commonplace. We've alreaqdy seen fairly low level 'mashups' but I know of several companies working hard on this one. If they succeed, then the door is wide open.
You're right about the one month's notice thing but the fact saas vendors are prepared to make that offer speaks volumes to me about their service commitment and confidence in what they have to deliver.
David and I are working on finding good examples beyond the 'usual suspects' because we know that's where the interesting stuff happens. I also believe that 'us believers' need to up our game in terms of explaining the value proposition.
I thkn you'll find that over the coming weeks, there will be some startling announcements and research findings that will help us all get a bertter understanding of what's really going on.
Users v Customers
Thank you all for this article, which contains a lot of interesting information/opinion in a small number of words.
I'm interested in the size of the market and I notice that all of the figures quoted in the article are numbers of users, not number of customers.
Running an SaaS service myself, I'm aware that there can be a big difference between these two figures. You can have a single customer who has multiple user access, but more importantly, the user count typically includes everyone who signed up for a free trial on a whim and then cancelled.
Does anyone have any figures for numbers of customer accounts on any of these services?
Our service (www.thepayrollsite.co.uk) currently has about 1000 active, paying, customer accounts. The number of users is far more impressive, but is irrelevant to anyone considering using our service.
Actual customers, "not shelfware"
Hi Steven,
That's an excellent question. I hope some of the others volunteer their numbers, and we'll try and get some data for you from FreshBooks, Salesforce and the like, but I can answer for Twinfield.
We don't have a free sign up process direct from our website (yet), although we do offer free evaluation following an online session. We have over 22,000 users who are actually paying a monthly subscription, which excludes free active users under special arrangements, for example in colleges and universities who are using Twinfield as part of their accounting course. The system currently supports 45,000 individual companies (sets of books). The disparity is that some accounting practices use the system to handle their "internal" accounts on behalf of a particular customer and the end customer doesn't have direct user access, and just receives reports. Currently there are around 700 practices of varying sizes using Twinfield, the largest of which runs the product from 32 separate offices.
I hope that helps, and they'll be more to come.
David Terrar
Mailto:dterrar@twinfield.com
web: http://www.d2c.org.uk and http://www.twinfield.com
blog: http://biztwozero.com
Excellent question
Winweb has 118K registered and is cash positive. Salesforce.com is making money off its 600K+ users but it investing 50%+ revenue on sales/marketing.
Excellent answer, but
Hi Dennis,
I know that Winweb are in a good position financially, but the 118,000 registered users doesn't fully answer Steve's question. It would be good to know how many customers that equates to, and how many of those registered users are active, rather than registered, tried it and moved off to somethinge else.
David Terrar
Mailto:dterrar@twinfield.com
web: http://www.d2c.org.uk and http://www.twinfield.com
blog: http://biztwozero.com
I'm confused
Alastair,
You can't have your cake and eat it. First you criticise us for over egging the pudding in the comparison, but when somebody comes back with some facts, you say TCO is irrelevant. I think most people will want to compare costs as well as the advantages and disadvantages. There are things you can do with a SaaS solution you just can't do, or are impractical with an on-premise solution. If you add that to the economics, you can make a strong case.
In our case, if the prospect wants on-premise, then our solution isn't for them. Like several of the SaaS vendors, Twinfield is fully committed to this delivery model. We don't do anything else. Gartner quotes figures for the SaaS market growing dramatically over the next 5 years, but nobody is saying that it will replace on-premise. Others will offer a hybrid model or a choice of both. The customer can compare and contrast and take their choice. Open Source absolutely has a place, and we promote an open source, SaaS solution alongside Twinfield for CRM, SOP, POP, etc. for companies where that is appropriate.
David Terrar
Mailto:dt@d2c.org.uk
web: http://www.d2c.org.uk and http://www.twinfield.com
blog: http://biztwozero.com