Any business that has to communicate with customers, regardless of vertical or horizontal categorization will come to WebRTC sooner or later, with three questions at the core: How soon do I need WebRTC? How fast do I need to roll it out? How much service capacity do I need to support it in my business?
Take a look around the unified communications and call center spaces. Everyone is either delivering voice and video via WebRTC or working overtime to get WebRTC support on the checklist, either to make inquiring customers happy or just to simplify the delivery of real time communications (RTC) services. WebRTC is already a must-have in the UC and call center space -- for those areas, the answer to "How soon?" is anywhere from "Yesterday" to "Immediately within the next six months."
Businesses less dependent upon real-time communications as a core technology need may be able to get by with existing voice and video setups as a transitional measure and those that don't currently have RTC have more time to integrate it into customer-facing web sites and internal business processes.
Regardless of the timetable, everyone can started with WebRTC immediately by picking up a free account at one of the many providers delivering WebRTC as a Service. Offers typically start with 5 free users and enable companies to immediately start testing out ideas on adding voice, video, messaging and other features to existing applications. Most offer in-depth software development kits and example code that can be wrapped around (well, into) existing applications for quick and easy implementation, with both stock on-line browser support and mobile iOS/Android support.
Using a WebRTC service means you can start development rapidly and scale up as needed fairly quickly. For larger, more aggressive implementations the question becomes how fast and how large the service provider can scale up, which leads into the second question as to "How much service capacity do I need?"
Scaling can get complicated very quickly. Do you anticipate most of your users coming from one geographic area or from around the world? Will you have tens, hundreds, or potentially thousands or more real time media sessions going simultaneously, requiring lots of bandwidth and multiple services? The need to serve large numbers of users -- large defined as some multiple of 10 for most cases and 100 or more typically -- also triggers a requirement for rock-solid reliability, with multiple 9s of reliability and fault tolerance.
Anyone with a large project encompassing hundreds to thousands of users doing voice and video at the same time is likely to also ask the question of building dedicated WebRTC services instead of going through a third party. Most of the time, the answer of build vs. buy should be "buy" for a number of reasons. Building a truly robust infrastructure for RTC services requires a lot of capital and network expense, as well as design and engineering expertise to make sure things will work as planned even in cases that exceed normal anticipated use. Only the largest enterprises and service providers are willing to make a commitment of resources to build a WebRTC service capable of effectively serving thousands of users or more at a time. Smaller to mid-sized projects may be able to get by with in-house resources, but at some point reliably scaling requires multiple data centers and lots of broadband.
"How much" is also important in the initial selection of a RTC service provider. It's easier to start and scale with one provider than to start with one and then have to relocate to a larger firm on the fly, tempting fate and service interruptions. Nearly all providers provide a pay-as-you-model, with pricing per WebRTC service usage going down with volume.
I'm not an advocate for building a WebRTC service in house, unless there are some extraordinary requirements involved . End-to-end security is the only case where I could imagine a need for internal WebRTC infrastructure.