What happens when a SaaS or Hosting provider goes into administration or bust

What happens when a SaaS or Hosting provider...

Didn't find your answer?

When you run your software on premise, you are resonsible for all aspects of your system.
Backups, upgrades, patching etc.

With SaaS and hosting providers, you move some of this responsibility to the provider.

However, you also enter into a complex 'supply chain'

You are paying the SaaS provider, who in turn pays a hosting company who in turn might be reselling services from one of the larger providers.

What happens if any of these companies get into financual difficulty?
For example the SaaS provider has not paid their bill, or the server hosting company is not meeting its finance agreements.
Your data sits on a server that you do not own and you are probaly not part of the agreement in dispute.

What steps can you take to protect yourself from this problem?
I am sure it is a question that some of the SaaS providers have had to address.

Paul Kelly

Replies (29)

Please login or register to join the discussion.

avatar
By merlyn
29th Mar 2013 18:00

It does happen

When 2e2 got into trouble they asked the firms who provide SaaS services using their infastructure to bail them out -

http://www.cloudpro.co.uk/cloud-essentials/5264/2e2-administrator-asks-customers-%C2%A34k-datacentre-bailout

As a company you should always do your homework when looking at cloud service providers, which should include asking who they host with and making sure they are a solid company, you will of course pay a premium for this but it's money well spent.

You should also ensure that you take regular backups of all your data which you control, be that to local storage on your network, a 2nd cloud providers or even to removable media such as tape.

 

 

Thanks (0)
By Captainblack
29th Mar 2013 19:25

Levels
You should try to protect yourself on several levels.
1) be sure that you understand the contracts you are signing, what you are entitled to, etc.
2) do your homework on the company you are signing up with, are they solvent?
3) try to discover if third-parties such as data centre providers are involved and do your research on them too
4) regularly download your data but bear in mind he data may be of limited use without the software application

Beyond this you are in the lap of commercial gods!
Captain

Thanks (1)
avatar
By merlyn
29th Mar 2013 19:30

BCP

Good advice from the Captain.

You should plan for a major issue/outage with your cloud provider in your business continuity plan.

 

Thanks (0)
avatar
By paul.k2
29th Mar 2013 21:04

Buyer Beware

Guys all of your points highlight the core problem.

On premise you retain some control, but the reality is that many small companies will not be able to make these enquiries or even realise it is an issue of concern.

The Cloud has its place, but you must understand what you are getting into. Is this the next big scandal???

 

Thanks (0)
avatar
By merlyn
29th Mar 2013 21:16

The cloud..

The cloud if used properly can be of huge benefit to companies as it gives them access to enterprise level infastructures for very low costs.

However done incorrectly it can cause massive problems.

On premise has it's own set of issues as usually servers/data are held in locations not designed for such things and can suffer from fire/flood power issues etc.

Sadly companies who don't understand the potential issues often speak to service providers who don't tell the entire story and in most cases use a 3rd party to host their services.

Not all cloud providers are the same, so never base who you use purely on cost, do your homework as which firm you trust to hold all your applications/data is very important decision.

There are lots of other considerations such as data encryption, backups, data protection etc.  But it's far too late to be getting into all that ;-)

 

Thanks (0)
avatar
By Cube
30th Mar 2013 13:16

Discussed recently

This was discussed recently.

Although one could not disagree with CaptainBlack's points 1-3, they are, on the whole, largely not realistic IMHO ("it's all ok until it's not", and "the customer will be the last to know" apply here in spades) - how many of us are tech savvy AND 'Report and Accounts' savvy AND know the specific SaaS / Cloud industry AND have the time and inclination .... it's becoming vanishingly small. However, his point 4 is golden.

Not only should SaaS/Cloud services enable excellent easy backups, they should provide customers with a cut-down (crippleware) version of their server software, that would run on a local PC allowing downgraded access to data ( & searching old back-up's), to address:

(1) significant outages - reasons many.

(2) the inevitability of one of these Saas / Cloud companies going spectacularly bust. (many are offering services for nothing, driving down costs across the board. Extreme predatory pricing in an environment where the commercial entry barriers are low)

(3) massive data loss / corruption.

(4) the inevitability of some of their customers going bust and needing to be able to access their data, either themselves, or for tax authority inspection, for years into the future.

Many of the SaaS providers whose software has it's origins in the Free OpenSource arena, keep a 'enterprise' version (commercial, hosted, supported and fee paying) and an 'OpenSource' version (essentially free, but without the latest bells and whistles and requiring some IT skills to set-up and maintain on a local server).

I am not aware if this provision of a local 'crippleware'  alternative is widespread, but it should be, for the reasons stated above.

Thanks (0)
Replying to Kent accountant:
avatar
By merlyn
30th Mar 2013 13:19

cloud

Cube wrote:

This was discussed recently.

Although one could not disagree with CaptainBlack's points 1-3, they are, on the whole, largely not realistic IMHO ("it's all ok until it's not", and "the customer will be the last to know" apply here in spades) - how many of us are tech savvy AND 'Report and Accounts' savvy AND know the specific SaaS industry AND have the time and inclination .... it's becoming vanishingly small.

Put me down as one ;-)

Cube wrote:

However, his point 4 is golden.

Not only should SaaS/Cloud services enable excellent easy backups, they should provide customers with a cut-down (crippleware) version of their server software, that would run on a local PC allowing downgraded access to data ( & searching old back-up's), to address:

(1) significant outages - reasons many.

(2) the inevitability of one of these Saas / Cloud companies going spectacularly bust. (many are offering services for nothing, driving down costs across the board. Extreme predatory pricing in an environment where the commercial entry barriers are low)

(3) massive data loss / corruption.

(4) the inevitability of some of their customers going bust and needing to be able to access their data, either themselves, or for tax authority inspection, for years into the future.

Many of the SaaS providers whose software has it's origins in the Free OpenSource arena, keep a 'enterprise' version (commercial, hosted, supported and fee paying) and an 'OpenSource' version (essentially free, but without the latest bells and whistles and requiring some IT skills to set-up and maintain on a local server).

I am not aware if this provision of a local 'crippleware'  alternative is widespread, but it should be, for the reasons stated above.

If using Linux as the OS and open source for cloud services then offering a local version would be possible, however for any Microsoft products (office 365, exchange, sql etc.) Then it's simply not possible due to the licencing.

Of course if you use Exchange with an Outlook client all emails can be set to cache locally so if your provider suffers an outage then you at least have those.

Having an exchange server on-site which is a replica of the cloud with automatic failover should the provider have an issue is achievable, but costly.

With all disaster recovery and Bussiness Continuity planning it all comes down to how much a firm is willing to spend v's the potential risk.

Thanks (0)
By Captainblack
30th Mar 2013 13:18

Hi Cube

Cube

One of the services provided by my busines is to assist prospective purchasers to evaluate solution options, the pros and cons of local vs hosted arrangements, the suppliers, the risks and benefits. In this way we help them to choose solutions wisely and know what they are getting into and how to protect themselves against the worst. You are right though, many would not do this well, or at all, on their own.

I have yet to see a reduced-functionality locally installable softtware product provided as part of hosted application arrangement. The best I have seen is the ability to download the raw data in various forms but frankly it's not a lot of use to most people without the software application.

So in this solution arena my advice is "trust everyone but cut the cards".

Captain

Thanks (0)
avatar
By merlyn
30th Mar 2013 13:34

Done but it's expensive
I've worked on some large cloud projects where the client has some servers on site, usually exchange, SQL, citrix and a DC. These are used as a warm failover should the cloud provider have an issue.

However these rely on big data pipes (usually via MPLS with a cross at telehouse direct to the provider) due to the replication being done and would be out of most firms price range.

Thanks (0)
avatar
By Cube
30th Mar 2013 14:15

High Standard

These  forums are lucky to have at least two people who are genned-up and fall into my "vanishingly small" group (I definitely exclude myself).

 

What I am beginning to 'take away' from these various chats about Saas / Cloud is that if done properly, with adequate back-up (and restore capability), it's not the cheap option many assume it to be. Isn't this always the way!

Whilst we have the attention of some obviously very knowledgeable IT fellows, I have a question.

I am sure that the vast majority of people reading this thread are in the small business sphere - 2 to 15 persons and will have to be making decisions on this themselves (perhaps not wanting to, or not being able to justify the costs of knowledgeable consultants). Where is the opportunity for them offered by SaaS / Cloud?

 

Where are the easy 'wins' (low hanging fruit) for the small business?

 

Note - technical waffle to one side. Easy applications / suggestions of counterparties / ideas of costs & savings. The 'wins' - names and numbers?

Note on my note - bearing in mind the very serious failings of SaaS / Cloud that we have covered.

To get things started ...

I can immediately think of only one - local server with encrypted (truecrypt - free) backup to cloud (Rackspace). However, this is a bit of a cheat (by me) cobbled together from info given by Merlyn in the previous thread.

Thanks (0)
Replying to ireallyshouldknowthisbut:
avatar
By merlyn
30th Mar 2013 17:14

Cloud

Cube wrote:

These  forums are lucky to have at least two people who are genned-up and fall into my "vanishingly small" group (I definitely exclude myself).

 

What I am beginning to 'take away' from these various chats about Saas / Cloud is that if done properly, with adequate back-up (and restore capability), it's not the cheap option many assume it to be. Isn't this always the way!

Whilst we have the attention of some obviously very knowledgeable IT fellows, I have a question.

I am sure that the vast majority of people reading this thread are in the small business sphere - 2 to 15 persons and will have to be making decisions on this themselves (perhaps not wanting to, or not being able to justify the costs of knowledgeable consultants). Where is the opportunity for them offered by SaaS / Cloud?

 

Where are the easy 'wins' (low hanging fruit) for the small business?

You are correct, the cloud if done properly isn't cheap but what it does it move your IT costs from a capex to an opex. 

The cloud also isn't one product, it's a number of services including online backups and SaaS.

People who have to make decisions regarding the cloud should get some proper advice, they pay an accountant as they have skills they don't have so why shouldn't making such an important decision for their firm be given the same treatment ?  And by proper advice I don't mean asking their local PC repair shop! 

Cube wrote:

Note - technical waffle to one side. Easy applications / suggestions of counterparties / ideas of costs & savings. The 'wins' - names and numbers?

Note on my note - bearing in mind the very serious failings of SaaS / Cloud that we have covered.

To get things started ...

I can immediately think of only one - local server with encrypted (truecrypt - free) backup to cloud (Rackspace). However, this is a bit of a cheat (by me) cobbled together from info given by Merlyn in the previous thread.

The best way for a firm to set themselves up proper use cloud services very much depends on what exactly they require.

But as an example for a small company which needs the basics (email and file storage) without using a cloud provider they would have to use the traditional method of purchasing a server, sticking it in their office, making a backup to tape each night and hoping that their office didn't suffer any disasters.

Setup cost of this would be approx £2000 with the on going costs of IT support for their server when required.

With the cloud they could setup hosted email with a decent provider, purchase a NAS box for file storage and have this backed up to a cloud provider (encrypted of course).

Setup costs of this would be approx £200 with an ongoing cost depending on how much data and number of users, but £50 a month would be a rough guide for 8 users.

So with the cloud it would take 3 years to spend the same as the setup cost with the traditional method of purchasing a server, thats not including the on-going IT support/admin needed with having your own server.

From a DR point of view there is a much greater risk to a locally held server suffering a problem than a well established cloud provider going under, but in the example I gave above even if that did happen with no warning the firm could either then purchase their own server to restore onto or choose another provider and copy all emails/data to them.

 

 

Thanks (1)
By Captainblack
30th Mar 2013 14:27

A few possibilities to consider

Cube. Here are a few possibilities for you to consider:-

1) hosted email (Google, Microsoft, Hosted Exchange via other providers, etc). These are inexpensive products and you can easilly protect yourself by keeping local copies of emails in your own email client software (Outlook, Thunderbird, etc).

2) hosted Backup. You should always have offsite backups in any case but you can get inexpensive and convenient offsite backup facilities (too many to list but LiveDrive for example). Here you can protect yourself by having local copies of your information as well as that stored in the Cloud.

3) Document Sharing. Here again there are numerous providers with decent hosted document sharing offerings. Again, you can protect yourself by having local copies.

A key line of thinking should be that no one event (disaster) should be able to cause you or your business to be completely compromised. Example: If your office burns down and you have cloud backups you should survive (from the data loss anyway)! For each 'cloud' idea you consider make that the number one question.

A second good line of question to ask yourself and the supplier is: what will happen when you no longer need (or want to pay for) the service? Can you get your data back? Will it be in a useable form? If not, consider the implications of having to continue to pay for the service even if you don't really need or want it.

Some thoughts anyway.

Captain

If it was easy anyone could do it!

Thanks (1)
avatar
By paul.k2
31st Mar 2013 10:44

Do we need a commercial / legal solution?

If you do not mind I would like to concentrate the conversation on the subject of SaaS.

I think this will be the first experience of the ‘the Cloud’ (or  ‘the Cage’ as one of my wise cracking colleagues likes to call it) most small companies and their accountants will have.

I am not an accountant, but an IT Consultant who works very closely with accountants.
Buying on site solutions for clients and helping them to manage risk is part of my offering.

When we undertake any conversation about a new on premise solution we discuss risk and how the business will cope with disruption.
We are often dealing with the software supplier and can negotiate specific requirements.
We also try to encourage the client to have an ongoing service agreement with an infrastructure company.

1. SaaS gives the impression that these problems disappear, but from what I can see, they are simply brushed over.

When I posed my question, it assumed users would be paying for the Cloud / SaaS solution. If you are using a free service, on your own head be it.

2. In the small business SaaS world you are one of many.
Someone pointed out earlier that SaaS is a numbers game. Its a take it or leave it relationship.
As Cube quite rightly asked, how many people have the time or skill to carry out investigations and how many of them will even acknowledge the need to do it.

3. SaaS gives small companies access to a range of facilities on an OPEX basis. Given the current economic climate and the clear intention of many major suppliers to adopt the SaaS model we have to face facts, the SaaS genie is out of the bottle.
This brings me back to my original question, what do you do if your SaaS provider or one of their suppliers gets into difficulty?

4. I have not carried out an extensive review of SaaS solutions, but I am pretty certain that few if any will:

i.    Provide backup facilities to a local disk

Many will provide some type of export facility, but I suspect this is not a straight forward task. As such users will not do it regularly.
Data centre costs are based on a number of factors, one of them being bandwidth. If a large number of users start downloading a copy of their data every day / week, costs would probably increase.

ii.    On premise versions
iii.    Escrow agreements

5. I think it is fair to say that we all know the problems. In theory we seem to know how to mitigate the risk, but in reality can we. If  20 or 0 of your customers suddenly lost their SaaS solution, can you quickly identify an alternative solution and migrate your customers?

Software businesses fail, but the nature of SaaS seems to create an urgency and complexity unlike that required for an on premise solution.
We, the customers, are trying to create a solution to the suppliers problem or is that the pact that we make for functional, convenient and seemingly low cost solutions.

SaaS is being offered to us as viable alternative to an on premise solution. It implies a level service and continuity that on closer inspection does not appear to stand up.
We have raised the questions, I think its now up to the Cloud / SaaS providers to give us the answers or the reassurances that we need.

6. I would like to conclude by posing a question to the accountancy profession.
Bearing in mind the nature of the relationship Cloud / SaaS suppliers have with their customers, should accountants not insist they make some provision for limited service in the event of financial problems?

Auditors and insurance companies are taking greater interest in company's IT systems. Perhaps its now time for them to turn their attention to the problems presented by the SaaS / Cloud environment.

Thanks (0)
avatar
By merlyn
01st Apr 2013 09:13

openstack

All the SaaS providers I've worked with allow users access to their data via Openstack (or similar), so can easily be scheduled to download to local storage using any number of freeware apps and most providers will assist with setting this up.

 

How useful the flat files/database would be without the actual software depends on what applications are in use, and the licencing as to if you could run a copy locally.

But does provide the ability to move to another SaaS provider if one goes bust.

That's not to say there are no problems with SaaS as it does depend on a working internet connection (unless you use locally cached apps, but this costs more) so relying on ADSL with no SLA may not be suitable, but as a DR plan you just all head to an internet cafe or home if there is an issue.

Thanks (0)
Replying to Duggimon:
Avatar
By I'msorryIhaven'taclue
01st Apr 2013 11:15

PDF Heaven

merlyn wrote:

All the SaaS providers I've worked with allow users access to their data via Openstack (or similar)...

...How useful the flat files/database would be without the actual software depends on what applications are in use, and the licencing as to if you could run a copy locally.

Until we parted company with our cloud software providers, we took the precaution of making pdf printouts of each month's audit trail, ledgers, and other key reports. One has to be able to access the data up to six years after the event for HMRC, by which time any software to drive files and databases of failed cloud software suppliers would be long forgotten.

Thanks (0)
Avatar
By I'msorryIhaven'taclue
01st Apr 2013 11:01

Evaluate with Webcheck?

Back in the days of Ford Capris and dodgy haircuts clients would often question their key suppliers' financial stability, whereupon we would off to Companies House to obtain company reports with which to provide financial analyses.

It would be an interesting exercise if a bunch of us were each to fork out a quid each on Webcheck and do the same with cloud software companies.

Finances aside, there are of course other factors: three years ago when we went overboard for a particular cloud software we had a client who for years had outsourced their bookkeeping to us whose VAT return liability for the quarter happened to amount to circa £30k. The following quarter the client moved their bookkeeping in-house (one of the obvious perils of the cloud system from an accountant's perspective), whereupon their newly appointed bookkeeper ran a VAT report for the previous quarter and found the liability should have been £65k not £30k. Needless to say, we lost the client's accountancy and tax business shortly afterwards.

So what went wrong? Yours Truly spent many hours recreating the VAT return generation process until finally the cause was revealed: the VAT return process worked perfectly well, producing the correct liability of £65k, provided one jumped into it directly from the cloud software's main menu; if however one firstly ran the previous VAT quarter's report (as we had done) and then jumped from that into creating the current quarter's VAT report and VAT return then the errant £30k liability was output.

Rightly or wrongly, I attributed that miswire to the extra layer of software residing between the nuts and bolts accounting program and the user friendly interface - it's just something extra to go wrong!

Thanks (0)
avatar
By User deleted
01st Apr 2013 11:08

Usefulness of backup ...

@merlyn is quite right - what use is a data dump to you if you don't have the accompanying software. Having 10-100 tables, guid’s as primary keys, stored procedures etc. is almost impossible to unravel without the software –

Although one way replication (down - not up) could possibly be used to copy the data ... and then two way replication introduces the concept of disconnected data; worked on centrally or locally

As for migrating to another system - under current circumstances pretty much forget it

Nevertheless, this whole issue of portable data is a subject that has been addressed in the past and I believe the whole thing stalled

However, some of the historic suggestions included having standard (or identifying) data fields for entries and <tagged> data. This would allow data to be moved between systems relatively simply; at least the software could identify the fields.

Taking this one step further, why not have standard file structures, after all how difficult is it to define a Nominal/General Ledger, Bank Account etc. because they all contain broadly similar data irrespective of application.

Once this goal is achieved they it would be an infinitely simpler matter for customers to switch provider, or upscale/grade rather than being locked in by a single provider that may not suit their requirements; or they may have outgrown

If the issue of portable data is addressed then any risk of provider failure evaporates because backups & switching becomes simpler and no longer a problem

Have a look at

Discussions--> Cloud Computing For Accountants Discussion Group

https://www.accountingweb.co.uk/group/cloud-computing-accountants-discus...

https://www.accountingweb.co.uk/group/cloud-computing-accountants-discus...

https://www.accountingweb.co.uk/group/cloud-computing-accountants-discus...

Don't know what has happened with '.. The Cloud Industry Forum are just launching a Code of Practice ..'

Thanks (0)
avatar
By merlyn
01st Apr 2013 11:22

bespoke
@Imsorry...

If you use a bespoke cloud SaaS solution then hard copy/PDF's are you only real option as if that provider goes under it will be very hard to get any database backups into a useable format.

For smaller firms who don't have their own DBA's, I would advise on only using SaaS with a well established product such as Sage, Microsoft Nav etc.

As that way any backup can be loaded into another version of the same software (either locally or with another cloud provider) fairly easily.

Thanks (0)
avatar
By User deleted
02nd Apr 2013 09:53

Don't agree with not using Cloud/SaaS ...

@merlyn

Rather like saying you might have a car crash so don't drive or don’t fly because aeroplanes fall out of the sky. Frankly these events have a far more terminal effect (and greater incidence) than losing some data and yet we still drive and fly – so being a doom-monger is probably not the most constructive way forward

Yes the result of an outage (Provider failure, disk crash etc.) with SaaS can be severe but one needs to balance that against in actual incidence of this type of thing (risk %) and the benefits one forgoes by not adopting a Cloud system

However, to recommend Sage or M$ as alternatives is simply incorrect. Especially as Sage had such an inauspicious entry into the market in the first place with displaying the users password ‘in clear’ on their screen after you had logged in; also their refusal to engage in this market for 10 years makes Sage on of the least mature candidates around. M$ bought their product from a 3rd party to enter the maarket and position itself in the 'Great Plains' (also purchased from 3rd party) series

A true/proper SaaS application (multi-tenanted etc.) should have been designed & written from the ground up to optimise the structure and approach for the internet – there have been may threads on this in the past on Aweb (multi-tenancy, latency, architecture, db structure … etc.).

Therefore, any application taking this ‘proper’ SaaS approach would generally not be portable to another application from the same manufacturer, unless of course they have written a parallel desktop version. Bearing in mind Sage’s approach to development, barely managing to deliver a proper app in the first place this is hardly likely.

With this in mind they would not easily be ‘slotted’ into an existing applications which would probably be desktop – unless of course you are discussing the Citrix approach which is a whole different ball game; allowing legacy apps to be ported to the internet and users conned with an ‘up to date’ badge, when in fact they are nothing of the sort; just the software house involved does'nt want to commit to re-development when they can string out the life of existing legacy products to the nth degree

Also don’t forget that the whole SaaS approach is incredibly flexible and it is perfectly possible to have the application running on the internet but accessing a SQL Server/MySQL/Oracle database located anywhere on the globe - even–on your hard disk (not practical but easily achievable) – or in fact in a multi-tier (x-tier) system you can have each tier in a different location – not efficient, but definitely possible

Finally, it would be possible for SaaS suppliers to provide web-services (api’s) to backup all or part of any db in the form of any other suppliers structure (if they wished to do this – probably not) … and so on … the options and solutions are endless, which makes the SaaS such a powerful flexible medium

And I am afraid that to suggest reverting to Sage etc. is simply a retrograde step that does no-one any favours and certainly inhibts progress for the sake of a percieved risk that is so small in percetage terms that a great many will never ever be affected

So all in all SaaS is the way forward but just do your homework before selecting a provider

EDIT

One thing that has always surprised me is that no enterprising company has developed an api/system to backup by reading one providers data and exporting it in another providers format - i.e. read Xero convert to KashFlow

This would get around nearly all the problems of provider company failure (liquidation) etc. and furthermore, make customer switching from one provider to another a fairly simple business - preventing lock in

Thanks (0)
avatar
By merlyn
02nd Apr 2013 10:12

Agreed

If you read back through my previous posts I said that the cloud/SaaS can be a very good idea for companies and my comments were aimed at paul.2k's specific question of

"This brings me back to my original question, what do you do if your SaaS provider or one of their suppliers gets into difficulty?" 

Not do they, or the chances of it happening but assuming they will then what steps can be taken to ensure the disruption isn't the end of the world for a business.

I was just using Sage and Microsoft as an example of a standard'ish application which if you had proper backups could be loaded into another providers infastructure fairly easily as if you use a bespoke platform as trying to get backups from that into another product would be very time consuming.

As I have said a lot of times in this thread it really depends on what applications a company uses, if it's just flat excel/word etc. docs then having a local copy works really well, if they use an application with a database backend then it's a lot more difficult as any backup won't be much use if the data is bespoke to the SaaS product and the company which hosts it has gone under. 

Ok so Sage was a bad example as their SaaS offering for line 50 is just their normal product which is accessed using Citrix. 

Line 500 is slightly better as it can be setup to be load balanced across multiple locations and has a web interface for users to access the system, however having worked on a large Line 500 SaaS project it' it's just really their legacy system with a WI so would only be classed as a level 2 SaaS setup.

Microsoft however are moving NAV to a proper SaaS platform and are currently at level 3/4 of the SaaS maturity model.

Companies should adopt SaaS as it's a really good way of negating a lot of the issues with hosting servers yourself, however do your homework regarding the company and also their IT infastructure to ensure they are fit for purpose.

 

 

 

Thanks (0)
Gary Turner
By garyturner
02nd Apr 2013 11:05

A maturing market

There's a difference between the kinds of checks you'll need to make to validate a enterprise cloud provider and a large scale multi-tenant cloud provider (like Xero). The latter is somewhat simpler and therefore easier. 

I'm happy to add some commentary around that.

SasS vendor viability encompasses a number of measures, not least their financial liquidity and ability to continue trading where as with any important purchase, it pays to at least validate the financial health of a supplier before entrusting your mission critical operations to them.Beyond the trading fundamentals, the SaaS vendor marketplace will rationalise as first generation vendors begin to stall or fail. Likely points of failure (and therefore areas for you to drill into) would be:Failure to scale - pulling together a minimum viable product or service is (relatively) easy. Profitably scaling up operations around that minimum viable product is another thing entirely, and encompasses not only obvious areas like the code and architecture of the product as more and more customers add load, but also customer care and critically customer acquisition - sales & marketing. 

Q: Ask how many paying customers they have, and what the growth has been and if they're slowing down. If they're cagey, you should be cagey.
 Inability to predict (and therefore finance a remedy) when scale hot-spots will appear - Specifically around scaling up the code and architecture, it's unlikely that the architecture of version 1.0 of a SaaS product which was good enough for 1,000 customers will be anywhere near capable of handling the kind of load that 10,000 or 20,000 customers will bring without significant investment and re-engineering. To compound that, these costs are hidden until the weaknesses are exposed.

Q : How many concurrent users are hitting the system at peak times, what was it a year ago, how are they managing to sustain performance, how many people are in their development team, do they have a dedicated operations team?
 Inability to acquire customers - again, it's (relatively) easy to find a few early adopters without much by way of professional sales or marketing discipline.

Q: How big is their sales team - what size was is 2-3 years ago, how big will it be in a year?
 Inability to compete - if by virtue of point 2. your limited dev resource is deployed in just keeping your app up and running, then there's little chance you'll have time to add new functionality. This will compound point 3. Moreover, if your sales and marketing engine is still at a minimum viable operational state, it's not likely to be working anyway.

Q: How much development focus is invested on adding new features, how many functional releases have they put out in the last year versus how much unseen architectural reworking?
 What do they expect their customer growth to be in a year or two years?

Q: Are you happy with current growth rates (therefore possibly a lifestyle business) or are they pushing aggressively (so, are they sufficiently funded and staffed to support that? Prove it?)

These are the top of mind things I'd ask.

Gary Turner
Managing Director, Xero
@garyturner

 

Thanks (0)
Avatar
By I'msorryIhaven'taclue
02nd Apr 2013 11:27

Disruptive Innovation

Speaking for myself, and from an accountant's perspective, I believe the next market leader will be the cloud accounting software that:

1. Provides usable backups capable of being imported into other software, preferably a stand alone, downloadable, hard disk based version of its cloud software. After all, we wouldn't go all year without making USABLE Sage or QB backups.

2. Provides a comfortable exit strategy - like a hotel: "Thank you, goodbye, do come again" ; rather than behaving acrimoniously and charging carry-on fees to access existing data. After all, one is terminating a service, not a marriage.

3. Carries out extensive and visible testing of its products, preferably by involving its users. How about paying a minor reward to those who unearth major flaws?

Thanks (0)
Replying to SteveHa:
avatar
By RussellD
02nd Apr 2013 14:45

Online backups can be used offline

I&#39;msorryIhaven&#39;taclue wrote:

Speaking for myself, and from an accountant's perspective, I believe the next market leader will be the cloud accounting software that:

1. Provides usable backups capable of being imported into other software, preferably a stand alone, downloadable, hard disk based version of its cloud software. After all, we wouldn't go all year without making USABLE Sage or QB backups.

2. Provides a comfortable exit strategy - like a hotel: "Thank you, goodbye, do come again" ; rather than behaving acrimoniously and charging carry-on fees to access existing data. After all, one is terminating a service, not a marriage.

Online50 provide Sage 50  online and strongly encourage users to take full backups after any changes to the Sage 50 dataset. The backup can easily be moved onto the local machine.  The backup will work on a local copy of Sage 50 (and visa versa). Sadly you would have to purchase the local copy of Sage 50 software, but it is easily and widely available.

Regarding an exit strategy, our licensing and service are available on calendar monthly terms. Online50 is the only company authorised by Sage to provide Sage 50 licensing on a month by month basis. Given the backup and calendar monthly terms it is easy to leave Online50 both from a data and financial perspective - which is why so many stay.

Russell

PS I do work for Online50

Thanks (0)
Gary Turner
By garyturner
02nd Apr 2013 11:45

Not realistic

@I'msorry - I can totally see offering an after-life desktop version of a cloud product would theoretically reassure some, but it's just not realistic. If you're unconvinced that now is the right time to move to the cloud in the absence of safety net like that, then its more the case that you're not ready to move to the cloud than the cloud is not ready for you.

Backups capable of being imported to other software : in the absence of a financial records database standard, there would be huge dependency on cloud vendors maintaining database structures and mappings to individual products to provide this. The best you could hope for in the interim is a full export of your data and a third party who would then convert it. There are some services emerging now http://www.movemybooks.com who will extract convert data from desktop accounting products like Sage to Xero for example, and I imagine it's only a matter of time before they start converting from other cloud products also.

But because of the overhead of tracking different versions of different products, this would most likely always be a paid for service rather than a standard feature.

Gary Turner
Managing Director, Xero
@garyturner

 

 

 

Thanks (0)
avatar
By User deleted
02nd Apr 2013 12:36

A great many first offerings were simply not fit for purpose ...

@garyturner

‘..the SaaS vendor marketplace will rationalise as first generation vendors begin to stall or fail ..’

Probably the over-riding reason for dropout is funding and speed to market - not technical ability. Invariably most start-ups in this business originate on a shoestring and it is often a case of develop the application or market something that doesn’t exist – chicken & egg

In the early days technical ability probably played a lesser part than marketing & fund raising in version 1 of most SaaS products on the market today.

Scaling, identifying hot spots (bottlenecks) etc. should all have been in the original design goals and as market penetration increases these areas become a by-product of funding and therefore resources. Clearly an injection of £xx million helps in this respect

I could name a few market entrants who are still in the game and regarded as high profile, whose products were originally simply not fit for purpose.

They chose to come to market too early and ‘wing it’ by attempting to get critical mass quickly, therefore influencing their ability to raise funding. Often those who tried to achieve a ‘finished’, all singing/dancing product, but took longer to market their offering invariably had a greater probability of falling by the wayside, because although the end product was better, they rather missed the boat
 
As a result a great many of the original SaaS offering had fairly questionable first versions in a rush to market

As you quite rightly point out it is all down to marketing, but not necessarily ability – and in this respect one only has to look at the phenomenal success of Sage as an example of marketing or buying market share first, and technical ability second

Inevitably the passage of time broadens the gap between established and new entrants making it more difficult for new players to gain traction, although the fact that this sector is an internet businesses does help with a level playing field

Whilst you have always been very open about difficult questions, there are providers who have been extremely cagy when asked questions and have often avoided providing responses – numbered among these are 2/3 of the household names in this field

One final suggestion to include in to what to look for.

Look at the enhancements all the SaaS players have made over the last 2 or 3 years and ask yourself whether the new features should have been included as part of the basic product on first release.

If your answer is that it should form part of a basic product and ask why it was not included – then you will be able to weed out those whose marketing over-rode all other considerations.

They were may well have been one of companies who released ‘not fit for purpose’ first versions and if this approach was adopted once then .....

 

Thanks (0)
avatar
By paul.k2
02nd Apr 2013 16:11

Thank you Gary

Thank you for contributing to the discussion Gary.

I have no doubt that you and the other vendors will take a few knocks in the course of this discussion, but I think it is better that vendors are part of it.

The points that you make are all valid, but they are also points a mature and well established vendor would make. Once upon a time your company was a startup, with a credibility issue to address and you have done so successfully.

If SaaS is to develop it is important that we have a vibrant supply base and this means we will see many small and probaly niche players. I am working with one just now, they have a terrific product, but their customer base will be limied to a few hundred.

Dealing with a large company is no protection. The recent high street failures are not necessarily good examples, but anyone can fail. The advantage of doing business with a large company is that someone might buy out the conpany, its remaining assets and customer base.

So I come back to my original question.

I think the reality is that we will gravitate to the well established providers trusting that they to are working with well established companies.
If any SaaS provider had a silver bullet answer to my question, it would already be in their marketing material and they would be making it a key point of differentiation.

My conclusion is that SaaS requires us to have a level of confidence in our supplier that on premise does not.

Thanks (0)
Gary Turner
By garyturner
02nd Apr 2013 18:05

We're all learning

@paul.k2 - I know what you mean, but in my defence I have been saying more or less the same thing for some time, even when we were much smaller than we are today. :)

Also, my perspective is always changing and being informed by our experiences. The fact that an online vendor can't always predict the scale and timing of the emergence of architectural choke points comes from our experiences of having to re-engineer important parts of Xero a few times over the last few years. And the fact that until you hit these design glass ceilings you don't know they're there.

While the initial barriers to market entry as a vendor are considerably less acute in SaaS than with on-premise, there's a bunch of hidden ongoing cost that SaaS vendors bear as they scale up in customer numbers, and these costs were not traditionally something an on-premise vendor had to factor in. In some senses, building a scaling SaaS product involves a lot more development work than an equivalently successful on-premise product - and any evaluation of a SaaS product I guess should include that in the viability calculation.

Frankly I didn't have this kind of insight four years ago, I do now.

Gary Turner
Managing Director, Xero
@garyturner

Thanks (0)
avatar
By paul.k2
02nd Apr 2013 18:18

Clarification

Gary one of my comments could have been phrased better.

Apologies. Your comments were all valid regardless of how well established your company is.

Thanks (0)
avatar
By machon
11th Apr 2013 11:49

Secure storage

Coming to this thread a bit late, my view is no storage mechanism is totally reliable. So, I diversify how I store data. I do this by using Network Attached Storage (NAS), and off-the-shelf crate with two drives in RAID formation. If either drive dies, I'm covered. If the box dies, the data is safe on two drives and I can get another box to put them in. Then I back up the NAS to the cloud. I'm only at risk if the cloud supplier goes bust at the same time as my premises burn down. Rob

Thanks (0)