COMPETITION BETWEEN SOFTWARE-AS-A-SERVICE VENDORS Dan Ma - - PowerPoint PPT Presentation

COMPETITION BETWEEN SOFTWARE-AS-A-SERVICE VENDORS

Dan Ma Robert J. Kauffman Singapore Management University May 28, 2013

Software-as-a-Service (SaaS)

• An important component of Cloud computing
• Bundle software application and services
• Deliver the bundle through a network
• Charge based on per use/ per month (no

upfront charge)

• Use one “standardized” application to serve

many customers; known as the “multi-tenancy” structure

• Build long term business partner relationship

with customers

The SaaS Market

• SaaS delivery will amount to >13% of worldwide

software spending

• By 2015, about 24% of all software purchases

will be SaaS

• Global spending for SaaS: US$14.5B (2012) à

US$26.5B (2016)

• Big name players like IBM, SAP, Oracle,

Salesforce.com, FinancialForce, Intacct

• Competition exists in niche SaaS application

market

Want an SaaS CRM system?

VS. $70

/user/month

On Demand

What makes the SaaS competition special?

• The offering is a bundle:

(Software application, IT services)

• Software is experience goods & Multi-tenant structure
• Long term business partnership

Different functionalities Different quality levels

Horizontal differentiation Vertical differentiation Incomplete information Sampling period and learning process Switching

Research Questions

• What are the pricing strategy and service quality

choices a SaaS vendor should employ to perform well in competition?

• How much should a vendor’s applications and

services be differentiated?

• How should clients choose the appropriate

vendor when they face incomplete information on the vendors’ offerings and the potential risk of being locked in?

Related Works

• Competition with heterogeneous products (Shaked and

Sutton 1983…)

• Horizontally differentiation (in software functionalities)
• Vertically differentiation (in service qualities)
• The role of switching costs (Klemperer 1987, 1989, Farrell and

Shapiro 1988, Cabral 2012)

• Switching costs increase/reduce prices
• Switching costs increase/decrease competitiveness of the market
• Other SaaS business model issues
• SaaS contract design (Susarla et al., 2003, 2009)
• Software investment incentive (Choudhary 2007)
• The impacts on traditional software market (Balasubramanian et al.

2008, Fan et al. 2009)

The Model

• The Vendors:
• Duopoly SaaS H and L
• High and low service qualities with high and low service costs

respectively

• Software applications with different features (Salop circle model)
• Decide the usage-based price
• The Clients
• Preference to application features is different
• Willingness-to-pay for services is different: θh and θl
• Face incomplete information of the application
• Can sample one vendor and learn Fit Cost
• Have the flexibility of switching to another vendor but incur

SwitchingCost

Timeline for SaaS Vendor-Client Interactions

The two SaaS vendors post prices, pH and pL, simultaneously The client will not know its FitCost for any SaaS vendor. It must choose a vendor based on its expected utility The client will use its chosen vendor's SaaS application and learn its FitCost. The client must decide whether to switch to another

• vendor. If so, it will incur a

switching cost S.

Analysis – Clients’ Utility (I)

• Backward reduction method
• At time 1: switch or not switch

“Marginal switcher:” the client who is indifferent between switching

• r staying with his initial choice.

1) The marginal θj –type client of Vendor H is defined by dHj

*, where

dHj

* is given by: θj ⋅ qH – t ⋅ dHj * – pH = θj ⋅ qL – 0.25 ⋅ t – pL – S

2) The marginal θj –type client of Vendor L is defined by dLj

*, where

dLj

* is given by θj ⋅ qL – t ⋅ dLj * – pL = θj ⋅ qH – 0.25 ⋅ t – pH – S

Analysis – Clients’ Utility (II)

• At time 0: which vendor to try out
• Consider a θ

l -type user’s comparison:

If he chooses Vendor H: 2 ⋅ dHl

* ⋅ (θ l ⋅ qH - pH - 0.5⋅ dHl * ⋅ t) +(1 – 2 ⋅ dHl *) ⋅ [(θ l ⋅ qL - pL - 0.25 ⋅ t) –S]

If he chooses Vendor H: 2 ⋅ dLl

*⋅ (θ l ⋅ qL - pL - 0.5 ⋅ dLl * ⋅ t) +(1 - 2 ⋅ dLl *) ⋅ [ (θ l ⋅ qH – pH -0.25 ⋅ t) – S]

Analysis – Market Segmentation

Analysis – Vendors’ Profit Optimization

• The vendors assign pH and pL simultaneously to maximize

profit. Vendor H: (pH – cH ) ⋅ (2 ⋅ dHh

* + (1 – 2 ⋅ dLl * )) – I – (1 – 2 ⋅ dLl *) ⋅ I

Vendor L: (pL – cL ) ⋅ (2 ⋅ dLl

* + (1 – 2 ⋅ dHh * )) – I – (1 – 2 ⋅ dHh * ) ⋅ I

When SwitchingCost is HIGH

• Opportunistic Pricing Strategy is recommended when
• High cost efficiency of SaaS (∆ServiceCost/∆Quality<θl) è SaaS H

can drive L out of the market and serve all clients

• Low cost efficiency of SaaS (∆ServiceCost/∆Quality>θh) è SaaS L

can drive H out of the market and serve all clients

• Non-Confrontational Market Segmentation Strategy

is recommended when

• Medium cost efficiency of SaaS (θl<∆ServiceCost/∆Quality<θh)

è SaaS H and L must coexist in the market è Each serves segment of clients based on the level of service quality preference

• è No direct competition between the vendors

When SwitchingCost is LOW

• SwitchingCost only benefits one SaaS vendor but hurts

the other.

• Lock-in power belongs to one SaaS vendor only.
• High cost efficiency of SaaS (∆ServiceCost/∆Quality) è

SwitchingCost affects SaaS H positively (higher price and higher profit) and SaaS L negatively (lower price and lower profit)

• Low cost efficiency of SaaS (∆ServiceCost/∆Quality) è

SwitchingCost affects SaaS L positively (higher price and higher profit) and SaaS H negatively (lower price and lower profit)

Vendor’s Service Quality Strategy

• An SaaS vendor is able to employ appropriate quality

strategy to gain exclusive lock-in power.

• The concrete strategy depends on the cost function.

When ΔServiceCost Is a Convex Function of ΔQuality Given QualityL è SaaS H should choose a relatively low level

• f QualityH (small quality difference)

Given QualityH è SaaS L should choose a relatively low level

• f QualityL (large quality difference)
• The efficiency of this quality strategy decreases in

SwitchingCost and clients’ vertical differentiation level ∆θ

Implications

• Offer Clients Flexibility
• gives clients the flexibility of sampling a vendor's SaaS
• ffering
• a marketing strategy involving free trials for potential

clients

• from free trial à subsidized trial?
• substantial investments to increases the free trial

conversion rate?

• designs contracts with greater flexibility (right to opt out)

Implications

• Lock in Clients beneficially
• customers’ lock-in not only from data recovery
• knowledge lock-in
• client’s competitive advantage may be lost.
• from adverse lock-in to beneficial lock-in
• cooperative strategic alliance
• cooperate, co-invest, and co-customize (partially for big

customers)

Implication

• Leverage on firm-level cost efficiency (relative efficiency)
• New startups or established vendor?
• Pure SaaS or mixed software vendor?
• Best practice, strong management
• over- and under -investmentà inefficiencyà

competitive disadvantage.