Ken Berryman – Winning in Software: Success Factors for the Next Era
Traditional software success factors
revenue growth drives company valuation, product innovation enables growth, time to market captures growth (get your sales force out selling)

Emphasise customer acquisition over customer experience

Changes
Growth is constrained by macro & micro economic factors

Now we have single digit growth rates, less of a tail wind

Customers want improved experience, not just innovation

India & China are facts of life
Competition for software margins

These factors changed the formula for success

Consolidation / M&A a fact of life now

Options for success

1. Finding new models for innovation (not just incremental features and functionality added to your product). Software industry is great
a. Innovation in Business model (a better way of pricing, bundling, delivering service)
b. Innovation in Usability (Design and User experience)
2. Customer loyalty (are customers buying your software because they have to or because they want to and want to and would recommend your software to others)
a. Relationship from initial contact through lifetime renewal
i. Typically a sales rep is only paid on front end of process (initial sale) and someone else is paid based on back end. How do you incentivize staff to deliver customer experience over lifetime of relationship.
3. How do you optimize software supply chain? 3rd party sourcing
a. licensing code from 3rd parties
b. having offshore developers in addition to internal programmers / R&D)
c. Monetize internal company intellectual property like drug companies do (# of patents filed)
d. global opportunity for growth (global demand for software)
4. Organize for productivity (eg WalMart pushed the retail industry, it outpaced on productivity)
a. Have very effective internal collaboration
i. increasing interactions due to email, IM, etc. how do you manage this so information flows back & forth well
ii. do you have a critical mass of skills

–

CEOs Look Ahead
With stuff that companies view as mission critical, they prefer to own the product, ability to customize the product specific to their needs, etc.

Quickly deployable, solve a need quickly = software-as-a-service

3 things are critical:
1. Security
2. simplicity
3. low-cost delivery

Software industry has not had a great reputation in terms of simplicity

Emergence of new devices
- what are the constraints behind getting software on cell phones, PDA’s, etc in addition to PC’s. (software growth on other devices)
o need standardization for this to happen, like the same processor in most pc’s. need standardization from hardware to software level.
o Memory is increasing
o The display and text entry are limited

Customers want appropriate licensing models: simple, provide value, flexibility in model and how solutions are delivered, more efficiency, investment protection

Challenges: what are customers looking for that Software companies are not giving them?
- solutions for a mobile work environment (people all over the place, on different devices, different service providers, etc). need to be able to access things like email, calendar, company directory, etc

SMB market is attractive, but the economics are difficult. SaaS is a good way to reach this market.

Emerging countries are attractive too. SaaS is a good model to reach these customers that don’t have large corporate infrastructure. They also like the pay-as-you go nature of SaaS.

PC Penetration is low in emerging markets, but cell phone penetration is high. Opportunity for cell phone application developers. Cell phone becomes internet device for people.

Consumers are dissatisfied with many tech products. People want a product to do more than one thing well (cell phone that acts as PDA, device convergence & multipurpose devices, etc)

Functionality is great. But simplicity & usability are key.

As mobile devices proliferate, software will become disintermediated, on-demand from wherever you are. Deliver information wherever you are, 24×7. information on-demand.

–

SaaS Executives Chart a New Software Industry

Subscription based delivery model for technology and services around a certain business problem. Changing the economic value proposition for customers – allow them to get significant return on investment with minimal upfront cost. Fast implementation, cheaper total cost of ownership.

Can provide a complex business process to customers at an affordable price. Clients can manage P&L from a visibility point of view, but also from a revenue recognition and cash flow perspective.

Can’t get economies of scale unless the applications are built for true multi-tenancy, where you can share the infrastructure among customers.

Cost for software company to host software (people, datacenters, bandwidth, capital costs, data centers) is cheaper than providing tech support for installed products. Customers get up faster and it costs them less to get up and running.

You need to understand your customers and decide if on-demand is right for them or if they still want installed products. On-demand isn’t right for every software company, depends on who your customer is.

Recurring revenue model means you need a lot of customers to get to profitability. Need some runway time. Once you are profitable, it is a beautiful model.

Understand fundamental drivers of
Enterprise software vs. SaaS
- every time you want to make 1 change to the system, every customer has to be notified or logged off
- revenue is very predictable (slow and steady)
- will take more capital to build an SaaS company. Have to make all the investments upfront, slow and steady growth.

Don’t require any involvement of IT staff within a client company. Can put out multiple releases per year. Application development moves fast. Need commitment to QA and integrity of software. If there is a problem, the customer can’t fix the bug. It has to work when you roll it out or every customer will know there is a problem. There is no versioning of the software. Every customer has the same version.

QA is huge. Also, need incredibly high levels of customer service – real time support. Everyone is a part of support team.

Need scale.

Software companies used to be able to get customers pay up front to pay for software development (up front payments help fund start-up). Now, customers are expecting to pay for value as it is received. Customers want software to be a tool to solve a business problem.

It is difficult in an SaaS world to generate rapid growth in revenue and to fund development from that revenue growth. Either expenses have to grow in relation to revenue growth or you need to look for other sources of capital.

How will large software companies work with small ISV’s? Alliances?
- partnership strategies for SaaS companies are fundamentally different than traditional partnerships. Traditional on-premise solutions have huge complexities. An ecosystem grows up around complexity. SaaS Deployments are easy, zero implementation services, only services relate to BPO redesign, training, integration with other platforms. If the objective of the 3rd party is to get consulting revenue, there isn’t much basis for a relationship w/ SaaS company. Traditional SI’s are going by the way side. New partnerships are with people who have existing relationships into the customer base (a certain industry). How does partner add value? Align incentives to reinforce value creation. Customers love SaaS model. Losers will be those who have inertia, winners will be those who can adapt (companies that can become facilitators of entry of the new model).

Bringing a product to market at a price point that over the lifetime may be equivalent to enterprise license, but where payments are spread out. IBM and other big companies can introduce these products to market. Bring high quality products into customer organizations.

For resellers, the pitch to customers is not about tweaking the software. It’s about making a pitch about how you can improve a customers business.

No difference between large and small customers. Only difference is how you market to SMB’s or large companies and customer acquisition costs. Scaling up to service large customers is easier than the opposite (enterprise companies trying to go downstream). I can serve 2 totally different markets with same code base.

ADP in payroll, Salesforce.com in CRM

Catalyst for SaaS growth?
As power shifts from CIO to CFO and other front line business owners, there will be a rebirth of purchasing, and that purchasing will come from business people. CIOs deal with infrastructure issues. They are not the ones who create value and revenue growth for a business (and SaaS tools are for growth, doing things better, etc).
IT directors are going to do more and more strategically for the company. They will move resources to the right things (away from having 5 guys support an installed HR application, another 5 guys focused on supporting an installed CRM application, etc).
IBM rallying around their SaaS initiative. Gives credibility to SaaS. Pushing the SaaS message. Just like with Linux.

SaaS – the benefits all flow to the customer.

SaaS is not sold into IT departments. Sell to business units. IT says no to everything. IT organizations that are forward thinking vs. those that are trying to maintain an IT empire.

Criticism of SaaS is that one size fits all does not allow companies to differentiate from their competitors. Customers who have unique business models want to customize software to fit their needs. SaaS companies need to anticipate where people typically modify software.

Easy for customers to switch away from a SaaS provider. This makes service / support fundamental to the success of the business.

Enterprise software model – to generate new revenue from existing customers sell maintenance and support and upsells and cross sells. Push things customer doesn’t want. SaaS model – give customers exactly what they want because you don’t get paid more unless customers use it. No benefit to overshooting customers needs.

If someone requests a feature, it is probably something that other customers want as well.

SaaS applications are easy to use.

Integration with ERP systems?
XML API – to integrate real time into other applications.
Batch processing / integration manager (non-real time)
Web services / XML API

Pure SaaS = a single instance, multi-tenant application that is only hosted.

Build an application from the ground up that takes full advantage of the internet.

Democratize solutions. People who could never before afford these types of solutions can now afford them.

Compare the time frame of a typical enterprise install (“we can have you up and running in March”) with a SaaS rollout (“does next Wednesday work for you?”).

–
Enabling the SaaS Revolution

SaaS infrastructure companies (that help other companies deliver SaaS)
Hosting and managed services,

Transition from traditional to SaaS
- taking out traditional license sale
- revenues can’t come in as quickly
- software code – taking single instance code to multi-instance and multi-tenancy
- getting programmers to think in terms of ongoing serviceable applications versus new features to sell more licenses
- operational issues – deploying and delivering a functionality to a customer. That doesn’t change. Customer is buying a function to make them better. But SaaS vendor needs to ensure security, SaaS 70, user management, etc
- still need to buy capital assets but revenues come more slowly
- getting past the I don’t know what I don’t know phase is a challenge for software companies. Companies need to be educated on how to make the transition. Consulting companies need to educate executives
- impact on infrastructure is huge. Embrace the channel if you start by going after the SMB market, for instance (if your traditional software company serves large customers). Don’t just pay them a commission on a sale. Salesforce.com is the anti-channel.

The Channel
- create a partner eco-system
- VARs and local SI’s start reselling SaaS applications and play a roll in the ongoing management of that customer
- Channel to SMB has been the same for 20 years and SaaS will not change this.
- Partners can build a practice and add value around the software.
- Partners used to sell $100,000 license and now have to sell $100 per user per month. How does partner survive.
o Probably won’t build distribution through the same channels. Knowledge consultants versus technology consultants. KCs are building context around a vertical market.

SaaS can deliver functionality that has not been delivered to people in the past because of restrictions on the desktop, etc

Key is not about delivering the Service, the service allows you to get out there in the market. Key is writing an application that provides unique functionality or capabilities to organizations that turn them into better companies. Have to deliver functionality that offers returns to customer.

VAR disintermediation??
If SaaS takes off, everyone has to change and taylor their solutions on the same types of metrics. Hardware companies, service provider, VARs, all add value along the value chain. Every company will have to adopt to add value. VARs that will be successful will flip their model and focus on services (recurring revenue) or become more vertically focused (trusted advisor). VAR can sell training, consulting services on making decision, be the comfort blanket. VARs will start to aggregate different solutions to create unique vertical service provider type applications (new virtual service providers that bring together other people’s IP and deliver to a specific market).
TechData and Ingram will be disintermediated. They will get changed the most.

Dealing with Internal IT departments to convince them that ownership of the customer should be shifted from IT to the SaaS provider
- engage IT from the beginning, don’t avoid them.
- Have to get them comfortable with the model
- Convince them that you can become a partner, that you are not a threat and can address all their valid and legitimate concerns around security, redundancy and scalability. You can turn them into advocates instead of obstacles in the sales cycle.
- Don’t avoid IT in the sales cycle because they will pop up at the end and kill the sale.
- Referenceable customers help.
- Monthly Release cycles: planning, development, validation, deployment (allows SaaS vendor to be very reactive to customer feature needs). But IT wants to “test” every release. You have to educate them that this is not possible if you release software every month.
- Once IT is comfortable, they are glad.

SaaS company doesn’t make more money until it’s customer sells more and gets more customers.

How long before software industry derives over 50% of revenue from SaaS?
- service revenue is cumulative, so it will grow slowly. 2010 is an estimate.
- 10 years from now is another estimate. Too many established big companies that sell software to businesses that will have a hard time transitioning to SaaS. Shock of revenues in public markets. Reconfiguring engineering group to release monthly instead of in 9-18 month cycles, building a support team.
- 10 years. Complicated by hybrid applications. Components will be placed inside firewall.
- 10 years. In 5 years Microsoft will still be at zero. But many new software companies will start as SaaS.

Everyone should be a student of the economics of software – from both the customer and vendor perspective.

Ebay is a great SaaS company.

–
Adopting the SaaS Model
As the market shifts, the goal is to provide customers something that is simpler to use, simpler to maintain, more cost effective for them. Still need to know what makes a solution successful for a customer and how customers can use that solution internally to transform their business. Provide all of that in a hosted model.

Build hosted application from the ground up.

One of the reasons SaaS hasn’t taken off yet is the breadth of functionality

SaaS companies, when building from the ground up, have to determine what is core functionality (start by just getting the basics out there) that a customer wants and can use. Functionality takes a back seat to the model & simplicity. Add functionality over time. Big challenge is keeping it simple while adding functionality. Get the basics out there and as customers evolve, stay one step ahead of them in providing functionality.

For existing software companies, pick key components / functionality of your existing bloated & complex software package and rebuild them from the ground up. Pull out core pieces and build something new.

Cannibalization has to be embraced. How do we make money with SaaS? What drives business model? SaaS will take longer to penetrate market but strategically is a good long term move. If you don’t cannibalize yourself, someone else will. When you price on-demand, look at how customers amortize licensed software over 3 years. Price SaaS offering so that over 3 years, customer would pay about the same as their licensed software, which they pay for up front. If you want faster adoption, make it slightly less expensive over 3 years. If you want faster adoption, make it slightly more expensive.

Segmentation – do a functional analysis of both offerings (current installed application and new SaaS offering). Enterprise segment has different functional requirements than SMBs. SaaS enables software companies to reach SMBs much easier because of price point. Price point also offers reinforcement on segmentation. Appeal to enterprise customer by saying old solution will give you additional capabilities that don’t exist in SaaS solution.

Talk to customers – “if we did this, would you purchase it?”

On-demand model – all the data is centralized. Customer still owns the data. But you can keep introducing services to help customers analyze data. In a packaged software environment, you can’t process data in real time.

Don’t change the look and feel on customers all the time. For new functionality, you won’t notice the change. Have old customers go into admin to turn new functionality on. New customers will see new functionality. Try to minimize global interface changes. UI changes, performance changes, functionality changes. Gradual change that is not intrusive to customer.

Release management is very different with a single instance of code and thousands of users.

Verticals – eg, CRM package for auto industry, paper packaging industry, retail vertical, etc. Database schema is important. Data model is important. Preset choices for customers in specific industries. Treat verticals as “additional capabilities”. What is turned on and turned off. All is in one model and each vertical has different things turned on.

Multi-tenant architecture – everybody is on one big database. Only have one database to upgrade or monitor. Or code is shared but everyone customer has their own database. Share hardware stack and code running on stack. When one customer has 20,000 users, they should not share the same physical database as another customer that only has 50 users. Need to be able to split things out. All the code is the same, but it points to different physical data stores.

Cost –Economies of scale after 3 year start up phase. Start-up phase is a sunk cost. Over period of time, cost of acquiring a new or incremental customer will become a greater percentage of cost. You will spend more money on support and account management in an on-demand model because you need to keep customer happy to make money every month. In packaged sale, making the initial sale is the most important thing. You have to manage SaaS accounts in order to retain the relationship. In multi-tenant architecture (SaaS) P&L over time, cost per incremental customer to host should go down. You get to amortize and make more efficient use of hardware and personel services and operations across more and more customers. Then cost driver becomes servicing the customer. They call you directly, expecting to be serviced.

Month by month contracts or yearly contracts – pay by the month or pay by the year and get a discount. Service will keep the client, not the software. We have to earn your business every month.

Internal changes to be successful in SaaS
- cannibalization issue – need to tell people internally that SaaS is a long term strategic asset to the company and needs to be nurtured and introduced side by side with enterprise offering.
o Have one sales team for both installed and on-demand products. Compensation model should make them indifferent to which one they sell.
- pricing and how you deliver the service (pure subscription, month to month, SLA?) – licensing, pricing and F&A decisions
- How to market it and position it

Standardization, simplicity are drivers that will push acceptance.