To WAN or SD-WAN ? – The APAC CIO dilemma

SD-WAN has been in the news in the last few quarters .In simplified terms, an SD-WAN enables cost-efficient use of public Internet and private networks. An enterprise CIO can now configure a mix of network services easily e.g, a mix of inexpensive Internet links to achieve high-speed bandwidth for non-critical applications, while continuing to use private networks( MPLS/Ethernet)  for mission-critical applications. The biggest advantage for enterprise CIOs is the ease of deployment as SD-WAN eliminates the need for expensive network engineers at every location.

The ability to centrally manage thousands of edge devices is useful for highly distributed enterprise verticals such as Retail, Mining, Healthcare etc. 45% of the CIOs interviewed validated that Branch expansion is the biggest use case for them .( Source : Frost & Sullivan SD-WAN end user Survey results – North America)

4 biggest business benefits of SD-WAN seem to be:

• Reduced Time to market for Branch expansion
• Cost savings – at an aggregate level
• Application visibility at branch level
• Cloud connectivity – Depending on criticality and nature of applications, mix of MPLS /dedicated link for a mission critical hosted ERP application vs an internet link

So the business benefits of SD-WAN seem clear but then why APAC CIOs are not adopting SD-WAN in an aggressive manner as should have been.

Few hypothesis on why this is the case.( Disclaimer : I am wording these Hypothesis as these are based on my limited set of discussions with 8 CIOs and a few ecosystem players across Australia, Japan and ASEAN in the last quarter )

• Hypothesis 1 – For typical CIOs SD-WAN ROI/TCO justification is still not clear. SD-WAN opens up the discussion at a different business level and the CIOs do not have a justifiable metrics to measure the benefits, atleast as of yet.
• Hypothesis 2 – APAC is a MPLS/Ethernet refresh market. And the issue is confounded as the typical MPLS contracts in APAC are long ( 3 years and in some cases 5 years !).So CIOs are pretty comfortable with this model which has worked for the last decade and may not be too keen to force the disruption. Again Hypothesis 1 may also have a role here. Lack of proper TCO /ROI models to justify the Total cost split – Overlay vs Underlay and quantification of the benefits.
• Hypothesis 3 – Skill set issue – SD-WAN opens up the skill issue in a new way . APAC CIOs are just getting comfortable with the talent and how they can use the existing CCIEs/CCNEs/ other qualified engineers who are conversant in a particular style/nature of configuration. SD-WAN requires a new set of competencies which the Network engineers of today need to adapt fast.
• Vendor marketing ambush – The last 2 quarters have seen an aggressive pitch by some of the SD-WAN vendors. The enterprises may be confused with the Value proposition of SD-WAN in the first place and the differences in SD-WAN solutions in the market. The pitches are also creating a sort of unreasonable expectations on what SD-WAN means .e.g., in 2 of my conversations with the CIOs, the biggest advantage of SD-WAN according to them was that they do not need to worry about last mile infrastructure at the branches. This is not true. CIOs will still need last mile and SD-WAN no way will eliminate the last mile underlay requirement. But what SD-WAN will allow is to look at your last mile infrastructure at each branch holistically and rationalize/optimize the last mile requirements.

From an analyst perspective, SD-WAN looks like a no-brainer. But there is a Wait-n-Watch approach by most CIOs in the APAC market.

If you are an Enterprise CIO struggling with SD-WAN implementation decision, would love to hear from you. We plan to launch our Research Calendar for 2018 on the APAC – Future of the Networks as well. So would be great to share and see how this journey spans out.

Sources:

• Analysis of the SD-WAN market – http://www.frost.com/sublib/display-report.do?id=A314-00-94-00-00
• SD-WAN end user survey results – http://www.frost.com/sublib/display-report.do?id=9B0F-00-1F-00-00