Monet -- Field of Flowers and Windmills near Leiden
In my last post, I provided a minimum definition of true SaaS and described the potential benefits for both vendors and customers who adhere to this now well-established concept in enterprise software. But the potential benefits are only realized as vendors move from delivering the basics of true SaaS to what I call SaaS InFullBloom, so perhaps I should trademark that phrase. Unlike the minimalist definition of true SaaS, knowing what else goes into achieving SaaS InFullBloom depends on the nature of the applications, the target market, the scope of functionality to be delivered, the geography and industry-specific readiness to take advantage/access specific capabilities, and many more factors. That said, there really are a core of preferred architectural behaviors which, taken together, separate garden variety true SaaS from great SaaS, aka SaaS InFullBloom.
SaaS done well is all about great software architecture and development practices — as well as about great capabilities that are consumed easily by customers. True SaaS, with just multi-tenancy, became mere table stakes in 2011; SaaS done well is an important indicator of durable success for the vendor and their customers. SaaS done well is about delivering on all of the potential business benefits of true SaaS.
So what’s SaaS InFullBloom? Here’s my list, in no particular order, of those preferred architectural behaviors:
- Elegant, extensive, and fully effective-dated configuration capabilities, to include pre-release sandboxes, regression testing tools, plenty of configuration and feature guidance delivered through the software (see the next bullet for more on this point), and a heavy dose of user (i.e. manager, employee, non-employee worker, applicant, beneficiary — everybody!) insight about the many places where not only the customer’s administrators but also those many users will want to and/or need to adapth the software to how they choose to work. SaaS done well can meet your special needs and retain configurations through each upgrade. SaaS done well can be embedded with the configurations needed to support specific types of work and workers, the regulatory requirements of specific industries and geographies, and the unique business rules and practices of individual organizations, with those configurations done by properly trained business analysts rather than by developers.
- Interrogatory configuration, which takes prospects from their earliest interactions with the vendor right through their repeated implementation cycles as their selected vendor releases several major functional releases each year along with the more frequent releases of regulatory and similar updates, has the potential for changing fundamentally the cost of sales, the elapsed time to revenue, and the underlying dynamics of the whole ecosystem of systems consultants for those HRM SaaS vendors whose underlying architectures lend themselves to dynamic configuration and whose business model sees the promise and are willing to invest in these capabilities. What TurboTax has done for US tax filing — making it possible for normal people to navigate arguably the most arcane concoction of business rules ever created — is what interrogatory configuration is intended to do for unleashing the capabilities of enterprise software.
- Extensive and easily set up cross-tenant and cross-geography inheritance of every type of embedded intelligence, from the obvious regulatory rules to those proprietary materials like competency models and dictionaries, salary surveys and compensation planning guidance, which result from the partnering of HRM consultancies which have built those proprietary materials with HRM SaaS vendors who can delivery that material as embedded intelligence throughout the relevant processes at minimal additional cost whether client-specific or vendor-provided, improves service quality and business outcomes.
- Models-based/definitional applications development, which is the state of the art in building business applications software, means there’s no procedural code written to create individual applications or configurations. Instead, there’s a ton of metadata that’s shared cross-tenant and then configured further by tenant. This approach to development is quite difficult to achieve because it requires considerable upfront investment in the necessary foundation tools (which, to my knowledge and specifically for the HRM domain, cannot simply be bought to support the complexity of some of these other preferred behaviors). But it offers such an enormous potential for improvement in the cost/time/quality metrics of bringing new funcionality to market that this now bleeding edge approach to HRM SaaS may create just the edge needed for newer arrivals to blow past their more established competitors.
- Extensive and easily set up cross-tenant data aggregation — with the permission of all the relevant parties, for crowdsourcing, benchmarking, and many other forms of cross-tenant collaboration.
- Systemic effective-dating — of business data, business rules, embedded intelligence, applications and application foundations. This is extremely difficult to do when applications are written via traditional procedural languages and much easier when it can be built into the fabric of a models-based/definitional development platform. But the payoff to customers in such a date-laden domain as HRM is in the huge reduction in errors, work-arounds, manual/shadow/logging side systems, sticky notes, mass changes (which are really about correcting what should have been correct in the first place), and the ever popular but oh so cumbersome audit database.
- Fully automated retrospective, retroactive and prospective processing — critical but also very hard to do unless you’ve got effective-dating handled systemically and have used a calculation engine rather than procedural logic to deliver everything from payroll to leave accruals.
- Mobile service delivery — not just reformatting but rethinking HRM, especially talent management (TM), workforce management (WFM), and strategic HRM analytics/decision-making, for a mobile world.
- Social/Collaboration — not just about socializing existing transactions or processes but about rethinking HRM from a workforce collaboration and collaboration use case perspective.
- Global — what platform capabilities for what target markets and in what countries backed up by what go-to-market plans and feet on the street, to include VARs and other types of distribution relationships?
- Analytics including what types of actionable, embedded, and/or predictive analytics with what types of visualizations at point of sale.
- Gamification — remember the suggestion box and the contests for great business improvement ideas with prizes for the best of them? Remember when, under the general heading of recognition programs, the employee of the month was given a primo parking space and you got a plaque for great attendance? Well, gamification is all of that and more. Using the same collaboration infrastructure noted above, gamification capabilities let us record publicly kudos to co-workers, earn badges for everything from good attendance to delivering great customer service, run sales contests to push older merchandise, and recognize every type of desireable behavior — and by exception call attention to those who aren’t behaving/producing/attending/etc. Although many of us still care about the ka-ching of compensation, there’s a lot of potential motivation, learning, teamwork, and even fun when HRM processes are rethought in these terms, so it’s important to have the right platforms in place to make this easy and pervasive when turned on by use case.
- Embedded intelligence of all kinds including content, guidance, exogenous data, best practice processes and business rules.
- Integrated HCM — core ERP/HRMS and core TM are inextricably intertwined, both as to end-to-end HRM processes as well as in terms of shared object foundations, consistent user experiences and workflow protocols, and the deepest of architectural enablers, and most customers aren’t able/don’t want to be systems integrators in perpetuity.
- Integrated talent management — with deep process-based and event-triggered integration so that all the good work of talent management isn’t undone by poorly handled ripple effects.
- Integration tools — true SaaS, cloud-based capabilities for the inevitable, needed integrations beyond HRM and based on the same principles as all other configurations so that business analysts can develop and manage these integrations without programmer intervention.
- Complete and correct object model for the scope of delivered functionality — if I see one more talent management or broad HRM application without properly distinct concepts of job and position, I’m really going to name names.
There’s a lot more to it, but this at least gives you a feel for my thinking. My preferred architectural behaviors “starter kit” is going on 250 pages of very dense slides, and the object model “starter kit” is more than 2,000 pages, so it’s not a trivial task to come up with the right behaviors and object model for even a modest HRM SaaS InFullBloom application. So it’s no wonder that vendors who start by building a modest HRM application run into some serious head winds when they decide to broaden their functional footprint and extend their customer focus.
Other closely related posts include:
https://infullbloom.us/?p=2654 Eight Characteristics Of Effective HRM And HRM Delivery Systems 12-2-2011
https://infullbloom.us/?p=2463 Impressions From HR Technology Conference 2011 10-10-2011
[…] Naomi Bloom, one of the most respected writers in the space takes fake SaaS to task in “There’s SaaS and Then There’s True SaaS” […]
[…] designs. The best of today’s HRM software reflect these ideas in their foundations; it’s not SaaS InFullBloom if it […]
[…] architecture. When I consider this change, I’m reminded of Naomi Bloom’s quest towards True SaaS – YaaS within HCP represents a True SaaS version of the old Hybris stack (see Andrea Stubbe’s […]
[…] therefore, in the platform are such capabilities as UX design, workflow, process management, inheritance across and within tenants, metadata management, systemic effective-dating, collaboration, video creation and embedding, […]
[…] if it doesn’t meet the strict definitions of those terms. Naomi has her own definition of SaaS InFullBloom in which she spells out the criteria even more […]
[…] is a true multi-tenant SaaS solution that would even qualify for my friend Naomi Bloom’s rigid InFullBloom SaaS. In my opinion, it has very modern cloud architecture and the least technical debt of any of the […]
[…] advantage of constant updates because Dovetail’s architecture is true SaaS. I personally like Naomi’s ‘Saas InFullBloom’ article, andGartner’s Reference Model for Multitenancy provides a great visual aid to true […]
[…] I see several vendors moving and/or trying to move in the direction of true SaaS InFullBloom, complete with proper object models, increasingly modern architectures, and real vision around the […]
[…] I see several vendors moving and/or trying to move in the direction of true SaaS InFullBloom, complete with proper object models, increasingly modern architectures, and real vision around the […]
[…] Naomi Bloom, one of the most respected writers in the space takes fake SaaS to task in “There’s SaaS and Then There’s True SaaS” […]
Naomi – great post and agree that True SaaS is the most efficient way to address these concerns for an ideal product feature set today AND tomorrow. Let’s not forget that the “ideal” feature set and capabilities were not necessarily priorities in the past. SaaS, above all else, allows rapid response to market changes. Weekly deployments and updates can expose new features for customers and this is made possible because the vendor takes on all of the effort of the upgrade.
Because there are no client resources involved in product upgrades, the bottleneck that exists with on-premise products is eliminated in the cloud. Now the feature are “just there” with the option to take advantage or not. The more frequently the better as far as customers are concerned because this gets issues addressed more quickly.
Without a true SaaS solution, vendors cannot afford to rapidly respond to requests via their products and will quickly fall behind the curve of a cloud empowered user base.
[…] many of these points may also apply to other true SaaS vendors, especially those which are fairly InFullBloom. So here I go, jumping into the fray, last but I hope not […]
[…] June, 2010, just when Al and Tipper Gore had announced their separation mid-2010, at a time when true SaaS InFullBloom was in its infancy. I was spending a ton of time with a wide range of vendors, some licensees of […]
Great post summarizing SaaS inFullBloom and you should trademark it. I have not come across such a complete list of feature set for a SaaS platform. I had compiled a similar list sometime back . One thing I can see missing is a PaaS services to go with the SaaS service.
No matter ho well the system is configurable there will be always some functionality missing that a customer would develop themselves and integrate with the SaaS using some sort of API. These extensions are most likely to be custom code and it needs to be hosted somewhere. They could use one of the public PaaS vendor but it then adds one more layer to the complexity. A custom PaaS platform to go with a SaaS should be another beneficial feature.
@sanskamat
Sandeep, You’re making a very good point about PaaS. For vendors who are primarily applications vendors, especially within the HRM enterprise software space, they definitely need to provide for product extension, ideally via an extensibility framework which is definitional in nature and builds up/uses the delivered object model, rules engine, workflow engine, and all the other infrastructural components. Whether or not they decide to go the distance and release a full-blown PaaS, which can be used to build entirely separate applications rather than just non-disruptive extensions to the delivered applications is a very large discussion. And it’s one to which I’ll be turning my attention shortly, always being careful of NDAs.
[…] an integrated, already multi-national if not yet global, true SaaS HRMS/TM platform with some SaaS InFullBloom capabilities that’s been built (or is currently being built) religiously with Microsoft’s most […]
[…] much more in this very big (but then they always are) release. But I’d like to highlight some things that struck me which may not be covered as much […]
One of your more mind-blowing posts Naomi. Maybe that’s what @Devasena was trying to say. 😉
On topic of interrogatory configuration, one of the bigger challenges is being able to enter that process at any point. When viewed as part of the life-cycle of the system a (re)configuration phase can be entered as a result of several triggers: regulatory changes, business environment changes, or regularly scheduled system reviews.
The last trigger involves assessing the functional and technical effectiveness of the system from a user perspective, support requirements (think analyzing tickets), technical performance, as well as a quick scan to determine if there are any new software products that deserve further investigation. Depending on the outcome of the assessment, there may be several entry points into the configuration step.
Definitely a lot of food for thought here!
You’re right about this needing to be a “re-entrant” process, and that’s where quite of a few of the other needed behaviors come into play, e.g. fully effective-dated underpinnings and applicability/eligibility. And yes, it’s a pretty content-intensive post. But I’m trying hard to mix up my posts between those which are more and less intense. Thanks again for your comments.
[…] Naomi Bloom, one of the most respected writers in the space takes fake SaaS to task in “There’s SaaS and Then There’s True SaaS” […]
[…] There’s True SaaS and, then there’s SaaS InFullBloomUS by Naomi Bloom (@InFullBloomUS) […]
Naomi, I agree with your general definition of what makes SaaS and with your proposition that HR is headed in the SaaS direction. I strongly agree that a multi-client, multi-tenant solution is necessary in order to realize the economies of scale that SaaS provides. We couldn’t run our 800 clients without a common data store and a single code line.
I would like to expand on the requirement for a strong object model. Of course the business objects such as Person, Position, Job, and client configurations need to be date effective and extendable using an object model. However, some of the greatest benefit is had by considering processes as objects and providing deep inheritance overrides. In this way, a client can use an existing process and override only those aspect of the process that are different for them. This is a vastly better way to provide configuration differences compared to the common practice of providing configuration options that end up making spaghetti out of simple processes. In many cases, it is a better approach than models since you start with an operational process. It also makes adaptation more rapid since it is not necessary to think of all the variations before constructing the process. It uses the power of the object model to make upgrades simpler and more reliable.
In the end, a SaaS operation can more competitive than older models if the object model is strong enough to be able to support widely varying requirements with a single code line.
Gary, I’m delighted to “hear your voice.” And I agree entirely with the concept of inheriting good practice process objects whose attributes and methods can then be configured, to include being overridden, for individual tenants or even components of that tenant. Where regulated functionality is involved, there may be necessary limits on the allowed configuration by tenant, but I’d always err on the side of robust configuration capabilities that leave their own effective-dated audit trail without requiring anything addtional to provide the needed audit capabilities. If not before, I’ll look forward to seeing you at HR Tech next Fall.
Great post, Naomi. I can’t believe that the question of effective dating (comprehensive, systematic effective dating) is still even an item up for discussion – it should be dial tone for any technology dealing with people and changes over time. Shame on anyone still not doing this right.
Each of these preferred behaviors are spot-on. Extensibility (integration capabilitiy) is so critical, especially the ability to go beyond HCM to the other business apps and data of the enterprise and beyond. Content is also critical (which you refer to as embedded intelligence.) I like to think of really good content as being essentially any object: people, data, video, discussion threads, goals, etc; all of which should be searchable and embeddable as appropriate. Social, mobile, analytics… In total it’s a daunting list for many I am sure, but a great signpost to success. Thanks for the post.
Yvette, Thanks so much for your encouragement. And believe me, I’d like to stop writing about systemic effective-dating too, but so very few vendors have handled this well and completely. I can really count them on a hand or two, which is pretty sad since we’ve been dealing with this since I wrote my first payroll application for an IBM 1401 in 1967. Between us Hussies, I’d love to hear from vendors as to how they match up to the capabilities described here rather than having me out them, but it may come to that. So vendors, if you’re reading this, we’d love to hear from you here or on the LinkedIn discussion at http://www.linkedin.com/groupItem?view=&gid=1772602&type=member&item=87503631&qid=f52935df-dcca-4a2d-9cca-959be4a42102&trk=group_most_popular-0-b-ttl&goback=%2Egmp_1772602
Great post, Naomi. This approach to SaaS certainly seems to make sense for large, complex, enterprise applications (I’m thinking Workday and the like). But for smaller, point solutions this list of requirements seems like added complexity, difficulty and cost.
On the customer side, there are some distinctions that need to be made in terms of requirements. My customer research (more than 60 detailed interviews in the last four months) shows that HR technology clients don’t always want a big system and in many cases prefer a simpler and more affordable approach. Depends on size and maturity of the organization.
Alex, Of course one parses the HRM domain object model to those object class structures and relationship, as well as to those attributes and methods, that are appropriate for the functional scope of what a specific vendor or custom development project will deliver. But that in no way reduces the importance that the underlying object model be correct and complete for that scope. With respect to the other preferred behaviors, which is by no means an exhaustive list but rather some important highlights, here too you must adapt what you’re doing to the scope of product or custom development. That said, good architecture is good architecture, and that in no way means that there’s added complexity from the customer perspective. Many tax payers use TurboTax without having a clue how complex is the interrogatory configuration that goes on behind the scenes. Rather, that level of smarts built into the system makes it possible for ordinary folks to do rather complex tax returns. And incorporating systemic effective-dating need not complicate use of the software. Rather, it makes it much easier for the lay user to get what they expect rather than expecting them to do workarounds and manual interventions — or to have an HR or other admin person do that for them. So, I respectfully disagree that what’s described here only applies to “large, complex, enterprise applications.” Not every nuance is applicable to every product — of course not — but most of them are applicable to a very wide range of point solutions, niche applications and such — in addition to being broadly applicable as the scope of functionality increases. All by way of saying that I hope your firm is producing SaaS InFullBloom behind the scenes and completely intuitive and productive software from the customer’s perspective.
good post, but little too high to understand Naomi
Devasena, Sorry you had trouble understanding the content of this post. If you have specific questions, you may want to post them here or join the discussion on this post over at the LInkedIn discussion group for the HR Technology Conference.