We’re all familiar with how the Space Race in the ‘60s and ‘70s led to many unintended technology breakthroughs for a range of non-related industries, including satellite TV, textiles, and transport. There are more recent examples of technologies jumping across industries, such as blockchain, artificial intelligence, and fiber optics.
Enterprise communications has benefited from a plethora of technology breakthroughs over the past decade. Modern communications would be quite unrecognizable to someone from 35 years ago, where the only means of communication was a desk phone connected to an analog PBX. Ironically, UC migrations would be the one area of today’s enterprise communications lifecycle that would be completely recognizable to someone from the mid-1980s and, of course, the Excel spreadsheet that goes along with migrations (actually, Lotus 123 - quite revolutionary at the time!).
However, UC migrations are about to change. Big data analytics technology is being adopted by enterprise communications in several areas, including migrations. I predict that, just as Lotus 123 was over-taken by Excel, so too will macro-driven spreadsheets finally disappear from UC migration projects and be replaced by “ETL” utilizing technologies developed for big data. Let me tell you how and why.
Big data ETL technology (i.e. extract/transform/load) was developed to process large and varied data sets typically generated from the Internet. It’s used to uncover hidden patterns, unknown correlations, trends, and preferences in order to help organizations make better business decisions. Sounds familiar, doesn’t it? Sounds a little like the hidden patterns of legacy PBX data that need to be identified and transformed into user-centric, advanced unified communications.
In the enterprise communication industry, we are facing some major trends that are relevant to this discussion:
1. The level of complexity (and volume of data) for enterprise communications is increasing
- Modern unified communications means multiple applications (voice, messaging, presence/IM, voice/Web/video conferencing, workstream collaboration, document sharing, etc.), multiple devices and numbers (desk, soft, mobile, personal, home office, etc.), multiple features (voice alone has more than 200 features) and multiple settings and preferences for each feature.
- The data associated with a single user can run to more than 1,000 elements in more sophisticated UC platforms, so it’s no surprise that we are talking about leveraging big data for a medium to large enterprise.
- Advanced dial plan (PBX and aggregation layer, on-net dialing, redundant connectivity, etc.) and hybrid cloud (i.e. various services that are delivered from third-party cloud providers) and fixed-mobile convergence, all adds to the complexity during a UC migration.
2. There’s been significant growth in the cloud communications (UCaaS) market, leading to an increased focus on UC migrations for enterprises moving to the cloud
- There are any number of market research reports showing that the adoption of cloud communications is growing at 20% to 30% per year. In fact, Gartner, in its recent UCaaS Magic Quadrant report, predicted that by 2021, 90% of IT leaders will not purchase any new premises-based UC infrastructure.
- Historically, the cost and complexity of migrating to UCaaS was seldom talked about as an issue. As the market has grown, however, the industry is now waking up to the pain and cost of moving to cloud communications. UCaaS migrations are not a smooth process; you need to plan and prepare well if you want to avoid loss of productivity and cost over-runs.
What’re the issues with the current UCaaS migration process?
The traditional collection of user/phone data -- manually through site surveys -- is error-prone and time-consuming (i.e. multi-visits and multi-iterations). Then, transforming the legacy user/phone data into UC-ready settings by way of spreadsheet macros is similarly open to human error and extremely difficult to validate due to the large file size and single-point access. It’s also difficult to accelerate migration projects, as the only way to achieve this is by throwing more people at the problem; and the “too many cooks” phenomena leads to even more errors and higher risk (and cost).
Also, given the time and effort required to collect and massage the data, it’s often several weeks before the data is loaded, and sites cut over, which means that the data currency declines (i.e. new users have been added or modified). To avoid this data currency issue, the customer is forced to implement a freeze on any changes for many weeks, which can impact their day-to-day operations.
Manual UC migrations typically result in 8% to 9% error rates. This can have a major impact on staff productivity when users try to log into their new cloud-based UC applications (not to mention the support staff costs to rectify these issues).
There really is no way to avoid these issues with the traditional migration processes. This is why the enterprise communications industry is starting to pay attention to big data and automation.
Over the years, many systems integration services companies have tried to change this model to avoid the costly and error-prone site survey process, but it has always proved too difficult and time consuming to manually extract legacy PBX data. There was no easy way to effectively extract, clean, normalize, transform, and validate this legacy PBX data, ahead of uploading it into the new UCaaS service provider’s platform, so companies were simply forced to select the “lesser of two evils” – site surveys and spreadsheets.
How is big data analytics able to improve the UC migration process?
Big data analytics ETL technology is highly suited to unified communications migration processes, as it removes the need for manual, site survey, and macro-driven spreadsheet data manipulation:
- Automatically discover and extract the legacy user/phone data (based on standard vendor templates and key design criteria) to ensure that only the necessary, live, phone settings are retrieved (i.e. avoids a single “dump” of data from the legacy PBX)
- Clean and normalize the data against multiple secondary data feeds (e.g. accurate user/number data from Active Directory) that can be automatically cross-referenced with the PBX data; collating messaging data; etc.
- Upgrade the data using big data “kettle” transforms (i.e. applying business logic, based on business criteria) to re-create the advanced UC-ready data automatically
- Validate the data before and during loading into the UCaaS platform (again, applying vendor templates to the data)
- Automatically test the newly loaded platform to ensure that calls can be made successfully
This migration automation for enterprise communications is revolutionary, and now possible through the integration of big data analytics with the latest UC fulfillment automation tools. And most importantly, migration data can easily be processed in batches so a highly repetitive process can be accelerated without needing to add extra resources; you just increase the batch size of the proven data automation process.
The key impacts of big data ETL migration automation on your migration process to the cloud include:
- Faster migrations (less effort for IT teams to prepare your organization for the migration, plus accelerated time to revenue for providers)
- Easier (smaller migration teams for both your enterprise and provider)
- Far fewer errors (human errors and data currency issues are minimized, resulting in typically less than 1% error rates)
- Much less risk of disruption to services post-migration
- Mass design changes are also now possible (e.g. not a like-for-like migration, but rather use the migration as an opportunity to significantly improve the underlying design by doing things such as restructuring naming conventions, re-designing dial plans, etc.). This is a major advantage for both end customer and provider and yet it’s almost effortless once the automation process is in place.
The resulting benefits from automating the UCaaS migration include:
- Much lower cost (real-world comparisons show cost savings of around 50% are possible with big data automation)
- Much less impact on end-users (significantly lower error rates -- 1% versus 8% -- and not requiring users to re-populate their personalized settings)
- Higher adoption rates (Nemertes Research has measured 30% higher adoption rates of new UC applications when automation is used)
- UCaaS providers can differentiate on migrations (as customers will achieve a much more positive ROI on the move to the cloud)
Let’s look at a recent case study
For a 15,000-seat financial services company’s migration to a cloud communications provider, the entire multi-office business was cut-over in a single weekend, rather than over many weeks. The total number of trouble tickets reported on Monday morning was 40. That is an error rate of less than 0.3%; something that would never have been achievable with a manual migration process.
UCaaS providers will very rarely mention the cost of migration to potential customers, but as you can see from the above discussion, migration automation is a major factor to be taken into consideration.