A software-as-a-service (SaaS) company delivers applications to customers over the internet as an ongoing service. SaaS companies provide and manage application access, security, availability, performance, upgrades, and maintenance. In the SaaS model, companies license software on a subscription or usage basis.
This is in contrast to the traditional software licensing and delivery model in which software is sold to the customer, installed locally, and maintained by the customer.
SaaS has grown due to the trends of remote work and BYOD, and end-user spending on SaaS is expected to grow 21.7% in 2022.
In this definition...
What are the features of SaaS applications?
As cloud applications, SaaS applications have different features and requirements than traditional, locally-hosted applications. SaaS features include:
Multitenant architecture: All users run on a single cloud instance with a common infrastructure and code base, but different customers’ data is still segregated.
Customization: SaaS applications tend to have some customizability out of the box, even for users with less technical expertise.
Mobility: Applications are centrally hosted and delivered via the internet. Users interface with the apps using a thin client, which could be a browser or application on a desktop computer or mobile device.
APIs: Application programming interfaces (APIs) allow organizations to integrate SaaS applications and create mashups, or applications that are combinations of multiple applications, a feat difficult to achieve with on-premises software.
Typical application services offered by SaaS companies include CRM, ERP, data management, human resource management, accounting and invoicing, ecommerce, project management, and web hosting.
Advantages of SaaS applications
An advantage of SaaS applications is that they tend to be more economical for customers than the traditional software model. Since there is no physical distribution of applications as they are delivered directly over a network, partners and middlemen are cut out of the sales process. Since the customer doesn’t have to manage the software on their end and applications run on thin clients, the burden of in-house IT expertise and infrastructure is lightened.
Furthermore, the subscription model decreases the upfront costs of migrating to SaaS applications, costs are spread out over time, pricing tiers can be flexible to match user needs, and freemium options are common.
In terms of convenience, SaaS applications provide customers customization out of the box, even with little technical expertise. Service-level agreements (SLAs) can be flexibly adjusted and capacity scaled as needed. Customers also don’t have to deal with the inconvenience or cost of major updates. Meanwhile, SaaS companies benefit from having only one version of codebase to maintain, allowing them to innovate more quickly.
Users can also work remotely on a large variety of networked devices, creating synergy with the BYOD trend. Users and vendors alike benefit from easier access management and data use monitoring.
The costs and risks of SaaS
While costs tend to even out over time, SaaS companies present the risk of acute disruption to critical business operations, security, and performance.
With uptime in the hands of a third party, users could experience interruptions to business functions when the SaaS company experiences downtime, software bugs, or other disruptions. If an SaaS company goes out of business, the SaaS service and data will be cut off. If the service is critical, the user will need to find a replacement vendor and may experience even more substantial losses.
While in theory, service delivery via the internet allows users to go mobile, a lack of coverage or loss of internet connection could result in loss of service. Thin clients and devices can also have compatibility issues that affect service.
The BYOD model also brings security vulnerabilities, requiring stricter access control protocols such as two-factor authentication (2FA). SaaS companies may not comply with security standards or be transparent about their security practices, data use, or access breaches.
Data ownership and use are determined by the SLA, but users may not have transparency into the actual data management practices of SaaS companies. Constitutional search and seizure warrant law does not protect all SaaS data. SaaS companies can encounter unexpected costs, risks, and data governance regulation conflicts, all of which can translate into service disruptions for their users and their users’ customers.
Performance can also be a sore point for users. Slow internet or poor connection can affect performance, as can latency due to geographic distance from servers. For organizations transmitting large quantities of data, internet bandwidth may be insufficient to meet their needs, making software and data hosting on internal servers and networks a better option.
While SaaS applications offer customization out of the box, large corporations may have greater overall customization requirements than multitenant architectures can offer. Customers have no control over versioning and can lose access to features they benefited from.
Pricing may not necessarily be competitive in comparison to traditional software licensing due to usage triggering unexpected changes in SLAs or price tiers, unforeseen training costs, downtime, loss of service due to SaaS company insolvency, and customer attrition over performance issues.
SaaS vs. IaaS vs. PaaS
SaaS, Infrastructure-as-a-Solution (IaaS), and Platform-as-a-Solution (PaaS) are all subsets of cloud computing and differ in the completeness of the product they offer.
SaaS: offers complete and fully-managed apps where the vendor handles all technical aspects of the service.
PaaS: is the next most complete. The model provides development platforms and other tools hosted by the vendor. These are used as a framework for the creation of custom applications by the customer’s in-house developers. The customer only manages applications and data use.
IaaS: provides only data center and computing resources. The company hosts infrastructure, including servers, storage, networking hardware, and virtualization resources. Customers must still manage data use, applications, and operating systems.