I got involved in a forum thread on LinkedIn over the last week, the gist of which was whether the implementation approach was different depending on whether you deployed CRM in the cloud or on-premise, and what the pro’s and con’s were of the two means of deployment.

I can’t recall having written on this before (which doesn’t mean I haven’t by the way!), so I thought I’d put together a post, particularly as I think there’s a myth or two that needs to be addressed.

If we go back to the early days of software as a service, as Salesforce.com took on the likes of Oracle and SAP, the debate seemed to polarise with Salesforce in one corner as web-based and easy to set up, and Oracle and SAP in the other, as client-server, complex, and inflexible.

The problem was (and to some extent remains) that these differences were often perceived as being down to the means of deployment (i.e. in the cloud or in-house), rather than simply being differences between the CRM applications themselves.

The fact that there were plenty of web-based, easy to set up, on-premise CRM options was often missed.

When I noted in the forum that I saw no great difference between the implementation approach for a cloud-based CRM project versus an on-premise one, there seemed some measure of surprise, which I suspect results from a lingering perception in some quarters that cloud is by definition quicker and easier to deploy.

There certainly are some aspects of cloud that do lend themselves to speed of deployment.  Firstly, there’s the potential to short-circuit the procurement process. One of the reasons for Salesforce.com’s initial success was that low up-front costs combined with user configurable functionality, and no requirement to install software locally, meant that purchasers could bypass the potential road-block of the IT team, and in many cases the procurement process altogether, because expenditure fell under sign-off limits.

Secondly, cloud removes the need to set up the software on servers as is needed for on-premise.

That said, if you’re looking to do something meaningful with CRM, then working with the IT department rather than around them is generally the preferred option, with most organisations now wise to the arrival of CRM in stealth-mode.

The software installation step might be a relatively big chunk of time if you’re looking at deploying out of the box functionality to a handful of users, but in larger projects it’s a comparatively trivial task in the overall implementation cycle.

It’s also worth pointing out that, contrary to some perceptions, cloud doesn’t have an inherent speed advantage in terms of the actual configuration and customisation of the system, with most modern CRM systems being quick and straightforward to tailor.

So from my perspective anyway, whether you’re electing to run CRM in the cloud or on-premise, has little bearing on the implementation approach itself. There’s still the need to define requirements, design and set up the system, integrate, migrate data, test, and get people using it in a consistent and structured way.

With a wealth of web-based, easy to set up and use, feature rich, CRM software options available in both cloud and on-premise flavours, the choice of whether to go cloud or on-premise should be dictated by individual circumstances. Cloud-based has a lot of advantages, which include:

  • Speed of deployment can be quicker, for the reasons I’ve described above
  • There’s generally less up front cost because licences tend to be paid for on a monthly or annual basis
  • You don’t have the cost and worry of setting up and running in-house servers
  • Updates and upgrades are quicker and easier

While these are strong benefits, there a number of reasons that for some organisations cloud-based might not be the right choice:

  • It requires reliable internet access. Relatively ubiquitous though it may be, not all would-be users have it in all locations that they work from.
  • There may be current or potential legal constraints about storing data remotely or in specific geographies
  • Where organisations are particularly sensitive about having direct control of the data they manage
  • In some instances (and this really needs to be analysed on a case by case basis) on-premise can be the cheaper option over the life of the system
  • Cloud users can be hostage to potential hikes in subscription fees
  • The impact of a CRM software provider going out of business is likely to be more acute for a cloud user rather than on-premise. I can only recall one example of this happening, but as the market consolidates over time I suspect we will see more casualties.
  • Finally, organisations running mission critical systems, particularly where there are high levels of customisation or integration, may prefer the ability to better control the management and timing of upgrades

While the majority of projects we work on these days are cloud-based, it’s certainly not all, and it’s important to keep an open mind and match the means of deployment to suit each situation. What’s right for one project may be a hundred percent wrong for another. But whichever approach you take, implementing it the right way is what determines success or failure, and that doesn’t change whether it’s running in a vendor’s data centre or down the corridor in your server room.

[Facebook] [Google] [LinkedIn] [Twitter] [Pinterest]

4 thoughts on “Cloud versus on-premise CRM”

  1. Cloud or on-premises also makes a difference because you end up talking to 2 different types of people. On-premises will get you going with integrators and the temptation to over-engineer and to over-customize might be more present (and maybe more pressing). Cloud is more likely to get you going with coaches and business consultants (who are more likely to have an independant take on what your CRM should be and should do). So I definitly vote for cloud 😉

  2. Good point about both types of CRM systems, Richard. In my opinion, a cloud-based CRM system is more appealing to SMBs because startups, small and medium companies could only benefit from an easy to deploy, maintain and update system. They are also more easier to try out, experiment with for the company to discover what they need and adapt it. At the same time, an on-premise CRM better suits bigger companies that know exactly what they need and they can customize it (actually, almost re-build it from the ground) and have the resources to maintain that CRM. This should be viewed on a case-by-case scenario, but cloud CRM systems seem to be a future of CRM.

  3. Richard, do you think there are any issues/points to consider for organisations with very large data sets? Is there a point when on-premise could be better for costs/scalability/integration?

    1. Hi Ivan – Good point. Larger data sets will generally incur additional storage charges, and, as on premise storage is generally pretty cheap, this would certainly be a factor in any cost projections. I’m not sure I see any insurmountable issues from a scalability or integration perspective, though that may depend on the platforms being considered. I’m just working on a project where there are very large data sets, and cloud is the likely route, so I will try and report some more concrete findings as we progress.

Leave a Reply

Your email address will not be published. Required fields are marked *