Skip to content

Thoughts on Saleforce

A community dedicated to making Salesforce, it's products, and partners better...

Click to register with for a FREE 30-DAY TRIAL!


Login Form

Lost Password?
No account yet? Register

Why is “No Software” Bad Positioning?

In commenting on my post entitled "Is Hosted CRM Viable?" Enough BS asks:

"I do not understand how you see local software as the answer to hosted software reliability problems. Could you elaborate?"

Simply put, local software could cache data locally and then replicate back to when available thereby allowing people to access and update their data when is offline. Not everyone would take advantage of this, but those who did could get past these downtime issues, and such a solution would scale a heck of a lot better than SalesForce currently does (By analogy, imagine if the all DNS queries had to route through one of the ten (10) root servers?!?) does have the Offline Edition, but it is only handles about 25% of what is needed.  For example, synchronization needs to be handled seemlessly and close to real time.  What’s more, there needs to be an option for something like Offline Edition to be the only interface users need, not the interface they use when the regular one is not available.  In other words, the user shouldn’t even need to know when SalesForce is down; everything would just keep working.

And local software can provide a lot of other benefits besides reliability. My opinion in this area is admittedly broader in scope.  If SaleForce continues to shun local software in their positioning, their competitors who do not will eventually bypass them.

For example, assume you want to update the address for ten (10) Contacts at the same company. Currently it is a very labor intensive projects.  It would be a lot easier if you have a grid of data that behaves like Excel than having to edit each individual record in a web form.  Of course, could add a web feature to make this use-case easier, but they will never be able to add enough features to keep up with the practically infinite data entry and data manipulation needs their users will have, and that local software can make easier.

6 Responses to “Why is “No Software” Bad Positioning?”

  • Lee responded:

    “It would be a lot easier if you have a grid of data that behaves like Excel than having to edit each individual record in a web form.”

    Er, how about excel?

  • Mike Schinkel responded:

    Yes, I’m aware of the tools for using Excel to work with data. However, they are not integrated into the process of using SalesForce on a day-to-day basis, they are non-supported open-source, they are difficult requiring significant technical skill (which I have but many sales people do not), and they are very “fragile” meaning you can easily cause errors or overright a large amount of good data with garbage.

    So yes, the current Excel tools exist and are better than having nothing, but as in the journey of a thousand miles they are only the first step.

  • Mike Schinkel's Miscellaneous Ramblings responded:

    Delphi for Sale! Wonder who will Buy?

  • Mike Schinkel's Miscellaneous Ramblings responded:

    Delphi for Sale! Wonder who will Buy?

  • Enough BS responded:

    Thanks for the reply, Mike. While this approach might be technically feasible, it isn’t without problems.

    This approach puts a burden on the subscriber to purchase and maintain adequate hardware and software to replicate the environment in which the host’s product runs. If the service providerís product is optimized to use Oracle as a back-end, for example, the subscriber would also need to run Oracle locally. This could be an issue in terms of cost, training, setup, etc. The same problem would exist with hardware resources, with purchase and maintenance cost being a major concern.

    The service provider could alleviate this problem for their customer by creating a more generic version of the product that could be hosted locally, and even offering installation and support for it. At that point, though, the provider would be maintaining essentially two sets of code for the same product. Not impossible, but perhaps not cost effective. A brand new provider might code from the ground up with this model in mind, and thereby only have one set of code to worry about. Still, the additional hardware costs may be enough to discourage many customers.

    It seems to me this type of hybrid product would also be a tough sell at most firms. If the subscriber has to host the product locally, they will ask themselves ďIs it really a reliable service? Why donít I just purchase the software and host it, and do away with the ongoing subscription costs?Ē They will expect a hosted service to have taken care of issues such as redundancy already. I can imagine the first several providers that tried to sell this approach would have an uphill battle in the market.

    Also, when a larger company moves towards a hosted provider, you can be sure it is with an eye towards cost savings. The built-in duplication costs of a hybrid approach could make this economically uncompetitive with a standard hosted model.

    Well, I canít see the future any more than anyone else can. I donít know where the hosted services industry will end up on the issue, but thanks for putting out your ideas, and providing the forum for discussion.

  • Mike Schinkel responded:

    Thanks for the detailed reply.

    I hear what you are saying, but I see it differently. I do see a problem with two separate versions, and I don’t advocate that. But I don’t see the problem you see with having the option to host a “local replication server” for those who need high availability.

    HA always costs more, and often the increase costs include implementation. You also can’t say “SalesForce should solve 100% of the problem” because they can’t guard against your connection to your internet Service provider going down, so a local caching solution really is a good solution, at least theoretically.

    And I don’t buy cost savings being the only reason for going to a hosted provider; it wasn’t my reason. Yes it can save in some cases, but I see many other benefits of going to a hosted provider not the least of which is minimal time and expertise required for many implementations. Another benefit is the ability to easily support users separated geographically.

    So what I was proposing would be, by my definition, brain-dead simple to implement locally. It could even be (optionally) delivered as an appliance.

    In another business I’ve previously tried using a home grown custom system, ACT 2005, Microsoft CRM, and they were all cheaper than paying’s annual fee (if you ignoring the cost of my time to build the custom system.) OTOH, the former two could not handle disparate geographies and the later would have been a nightmare to implement (and maintain) for my needs. So the thought of paying for hardware on which I could run a point-and-click install for a “local replication server” does not offend my sensibilities from a cost perspective at all.

    And remember, the local hardware need not be able to handle peek loads efficiently; it just needs to be able to handle them in the rare case that (or your ISP’s connection) is unavailable.

Add your own comment...