Pricing Seminar Speaker Notes April 2013

Pricing Models and Considerations in B2B and B2C SaaS

Dave Litwiller, Executive in Residence, Communitech

Seminar Given on Tuesday April 23, 2013 at the Communitech Hub, Kitchener, Ontario

Overview: Pricing models for B2B and B2C SaaS, including cost of capital considerations for early- and growth-stage vendors, profit engines, and working capital models

Three Main Factors which Influence Pricing: Profit Pools, Cost Structure, and Industry Stage of Development

- Profit poolso Industry

Market share At scale, the top two players typically command 65% of the revenue and

85% of the profit. Very different situation at a smaller size, and during the formative years for the business. Need to define your market in such a way as to be a clear #1 or #2.

Contemporary example, Q4/’12 smart phone sales:1

1 http://www.slideshare.net/bge20/2013-04-mobile-overview

©David J. Litwiller, 2013

Only about 9% of companies in a market sector will sustain outsized growth, and return above their cost of capital long term

2

Market chain In a market chain, there is typically a strong disproportion of where the

profitability lay and is moving, which confers where the pricing flexibility is Example of the PC Industry:

3

Lanchester dynamics – Competitive Dynamics

2 http://www.smi.ethz.ch/education/courses/corporatestrategy/Slides_AS2011/Bain_How_to_keep_on_growing_handout.pdf3 http://data.ciqol.com/upload/pdf/2011-01/110113042450_933316.pdf

©David J. Litwiller, 2013

To compete against a monopoly or duopoly head on with success requires marketing muscle 3* the leader

To compete in a defined but less consolidated market requires marketing muscle 1.7* the leader

Either of the above levels of spending are far out of reach for most start-ups in SaaS and mobile apps

Therefore, almost all SaaS start-ups are working to redefine their market with a new niche in which the insurgent can be a leader

Pricing thus needs to be strongly rooted in recasting and reinforcing a new segmentation where leadership is attainable

o Best is if the market is redefined in a way, and pricing with it, where the existing strengths of the leaders become liabilities through inhibition or inability to compete in new ways

The basis of redefining the market always comes down to one of three primary bases to drive new usage models or capabilities: better performance, lower cost, or better customer experience

Examples of how pricing reinforces cognitive coherence However, never lose sight of the overriding priority of fast execution speed

and rapid learning. Pricing is only a supporting tactic

©David J. Litwiller, 2013

Industry Stage of Development Smile Curve, Stan Shih

4

Internal Pricing Factors

- Customer base Whale Curve, Kaplan

Activity based costing Most profitable 20% of customers typically generate between 150% and

300% of total profits Middle 70% breakeven Least profitable 10% of customers can cost the vendor 50% to 200% Graphically:

4 http://www.google.ca/imgres?imgurl=http://linuxpundit.files.wordpress.com/2010/05/smile-curve1.jpg&imgrefurl=http://linuxpundit.wordpress.com/2010/05/26/will-android-drive-mobile-commodization/&h=580&w=600&sz=35&tbnid=uhbGK9RU8QZYbM:&tbnh=90&tbnw=93&prev=/search%3Fq%3D%2522smile%2Bcurve%2522%26tbm%3Disch%26tbo%3Du&zoom=1&q=%22smile+curve%22&usg=__9o12wjqof3bLrEzZBRnsFzudLBk=&docid=AYWHAzuIaUzVQM&hl=en&sa=X&ei=_4hmUfeHL6Kp2gX6roGwAw&ved=0CDgQ9QEwAw&dur=18647

©David J. Litwiller, 2013

5

Inferences: o Segmentation, know the real operating cost of servicing customers

at the tail end, and know what the game plan is to move them up to greater profitability through one form or another (premium) or reduced service levels

o If you’re going after a proven market, figure out how to leave the unprofitable customers with the incumbent vendors, and disproportionately appeal to the most profitable segment

5 Time-driven Activity-based Costing: A Simpler and More Powerful Path to ... By Robert Steven Kaplan, Steven R Anderson. Harvard Business Press, 2007

©David J. Litwiller, 2013

- Cost-based Pricing waterfall Hard and soft costs

6

o Costso Direct vs indirect costso Warranty, repair, maintenance programming – if too early to have predictive internal

metrics, rely on benchmarkso Bad debt – rely on benchmarkso Be mindful of discounts and discounting authority, especially as the sales team and its

distance from the founders grows

- Cost of capitalo Equivalence example – start with questiono Cost of capital at seed to series A is between 50% and 90%, meaning you need to provide an

annual return at that rate Need roughly 6% of the up-front selling price as a monthly subscription for a two

year deal, 5% for a three year, and 4% for a four year If you exchange an up-front payment for a monthly subscription at a lesser ratio,

you are substituting high cost capital (yours) for (generally) lower cost customer capital

6 The Price Advantage, Baker et al., McKinsey

©David J. Litwiller, 2013

o Source of Capital – Financial model Gross margin Working capital Investment capital

Pricing to Capture Value

Value - B2B – Large ticket items

- Learn all the aspects of how your product creates value- Structure early deals and instrument product to provide full business-level measures in the hands of

the end user for how value gets generated through lower costs and greater revenue in their context, as well as the related purchasing and deployment issues they face

- Document significantly- Aim to capture 50% of value add, be willing to settle for 30% of value add

o Further reference points: Cost to (re)create (look back) Future incremental income (look ahead) Market benchmarks (look around)

- As pricing negotiations progress, prevent value justification from blending into pricing. If a customer says he does not value something, if at all possible, take it off the table so that pricing is attached to explicit value. Don’t let people have something for free, and if it needs to be thrown in, then get a concession on something else.

- Alternatively: RoI B2C, B2B – Two year payback for private enterprises for the product to be able to substantially penetrate the mainstream, and 6 months to 9 months to get a really fast start. Longer paybacks are the nearly exclusive domain of government and near-government projects. Moreover, the relationship between payback and demand is non-linear.

©David J. Litwiller, 2013

- On the question of further value and price increases, above the minimum, whatever the baseline quantum of expenditure to get the targeted payback, larger increases in absolute cost have an appreciable impact on adoption, and more so as the size of the investment doubles and beyond

- One other thought: Whatever the business case says, the CFO is assuming the project will take 50% longer and cost 50% more than the initially expected outcome. Empirically, the CFO is right.7

7 http://www.zdnet.com/2013-erp-research-compelling-advice-for-the-cfo-7000011619/

©David J. Litwiller, 2013

Value – B2C

- For products used privately, with little or no social component, suggested reading: Inside Intuit- For products used socially, or those with social signaling value, suggested reading: Spent, Geoffrey

Moore- People are often much more influenced by anchoring, reference points, social proof and social

signaling than they are by more analytical factors

SaaS Financial Model Gearing

- SaaS: Know the CAC, churn, net CLTV (with impact of churn)o Scale if CAC is paid back in gross margin from MRR in less than a yearo Lifetime:

1-2 years for consumers 3-4 years for SMEs, or for a 50K SLoC system (new or modified LoC) 5-7 years for large enterprises, or for a 500K SLoC system

o If CAC is not paid back in a reasonable lifetime, the business is in slow motion suicide- Services: 2* loaded labour cost (benefits, travel, vacation, overhead time)

SaaS – To Freemium or not to Freemium, that is the question

- Distribution of SaaS w.r.t. Freemiumo About 40% of SaaS companies maintain a long-term freemium pricing modelo A much higher proportion of start-ups start out as freemum, but go on to realize that it

doesn’t work well for them. - Implicit bet:

o B2C - That there are 1% to 2% of users in who are hypermotivated to pay vastly more than it costs to deliver the service and embedded support to them.

o B2B – 2% to 10%, in a few extreme cases as high as 20% - A lot of orthodoxy about should one or shouldn’t one do freemium. There is no absolute right or

wrong. SaaS splits down the middle.- Freemium has to be a paying customer acquisition strategy. It is not a business model- Often, it is just wishful thinking: people use the product when it is free, but when you try to charge,

they resist or exit- Conditions to do Freemium:

o Very large marketo Incremental cost of service is very close to zeroo There is a compelling reason for users to refer others in to the service (drive up # of users)

©David J. Litwiller, 2013

o There is a clear segmentation of behaviour within the user base to drive an upgrade path to paid tiers; engagement inexorably drives people toward conversion (add clients, add files or pictures, then hit limits)

o The value of escalating to premium can be quickly and clearly explainedo There is sufficient funding to carry the business through substantial operating losses before

the onset of sufficient paying customerso People need it free to be willing to give it a try in large numberso The system benefits greatly from network effects and big data inference possibilities

- Gold standard for engagement to know that freemium is appropriate: Cohort analysis which shows that the proportion of users who upgrades increases with time, not dwindling after the first month or two, and, users who abandon the service miss it so much that they come back after time away8

- Guideline: If there is any premium segment willing to pay at all, there are at least two more ultra-premium segments willing to pay double and an order of magnitude more

o Heuristic: Extending the premium ladder from a single rung by one or two further rungs can double ARPU

o To avoid irritating loyal current customers, consider grandfathering them for price increases- Two flavours:

o Directly paido Indirect: access the user base (ads, subscriptions)

- Heuristic: It takes 2* more start-up capital to fund incubation of a freemium business vs. one that is not, since extensive investment in integration and onboarding (fabulous early user experience), delayed onset of revenue from premium tiers (often as much as two years for full premium upsell to occur), underlying economies of scale, and network effects/big data

- Overall: Be flexible when offering freemium and sensitive to whether the predictive analytics are really favourable. Some of the best moves start-ups have made have been to abandon freemium to either ad-supported or paid base tier SaaS.

Elasticity

- Price Elasticity, a.k.a. price-resistance or price-volume

- Price Elasticity=% change∈quantity demanded% change∈ price

- Maximum ratios as a starting pointo B2B, -1.8o B2C, -2.5 (discretionary)

- Typical is about one half those levels- Depends to a large extent on the size of the purchase and the extent of related purchase and

deployment obligations- Track over time

8 http://discussion.evernote.com/topic/22864-would-go-premium-if-these-features-were-offered/

©David J. Litwiller, 2013

- Early days issue: If in trouble getting sales in and you’re wondering if pricing is the issue: call up five leading prospects and ask them if they’d deploy immediately if you dropped the price to zero. More often than not, price is not the issue.

Pricing Strategies

- Demand-based:o Maximize exposure/penetrationo Skim

- Competition-based:o Conform for recognizability, or establish price leadership, or establish premium

- Cost-based:o Mark-up based

- Differentiate, often based on simplicity or fast configurability- Time-based:

o Opportunities for intentionally rapid time variability of pricing, to stimulate loss aversion behaviours such as panic buying or even hording

o Caution: getting customers used to waiting until ends of financial reporting periods, since they know that salespeople are anxious to make quotas and can extract large concessions on pricing and terms

- To Have a Bit of Fun: Temporarily withhold cost information from your marketing department, and ask them to establish what the selling price should be

- Pricing should be subject to A/B testing and rapid iteration, especially in the early days, just like any other significant aspect of competitive differentiation and durability

Two Sided Markets

- Know which side is the product (virality, network effects, big data, and supply or demand), and which is the payer

- Provide the subsidy to the more price sensitive or quality/convenience side of the marketplace- Processing fees generally range from 2% to 10%, with the higher end of the range belonging to

systems which get at previously landlocked supply- Considerably lower pricing prevails in financial instrument exchanges- Focus initially on a landmark deal, even if it is in a narrow niche (ex: Square-Starbucks); and

relatedly, don’t alienate marquee providers- Make sure the marginal cost of delivery is very low (no FreePC)- Network effects are acutely strong in two sided markets: market shares of the top players tend

toward 90%, 9% and 0.9%, vs. the weaker network effect norm of 40%, 25%, 15%

©David J. Litwiller, 2013

- Discount for high volume players in the market, because they provide the liquidity, social proof and draw players in

- Generally, charge low for basic listing, in order to maximize visible supply or demand. Charge more for completing a transaction or for premium services

- Advice: http://www.forbes.com/sites/ciocentral/2013/02/07/5-tips-for-building-a-two-sided-online-marketplace/

- http://pandodaily.com/2012/11/20/liquidity-hacking-how-to-build-a-two-sided-marketplace/ - Ebay: 4% to 10%9, AirBnB 10%, Amazon 12%, oDesk 10%- Other: Adobe, Q: what about PayPal?

Pioneer Customer Pricing

- Establish terms and framework for others to follow

- Fast start

- Target entity significant enough to influence others favourably in the competitive landscape

- Design influence

- Rapid, iterative deployment

- Mutual disclosure of issues and opportunities, organizational access

- Rights of publication and publicity

- Generally: Steep discounts on up-front pricing (perpetual license, professional services) , but retain a resemblance to arm’s length pricing for running fees (annual maintenance, subscription)

- Most favoured nation clauses: Exclude pioneer customer pricing

OEM

- Discounts, service, delayed onset of cash flow, but, recurring and sizeable follow-on volume

- Practical limits in software in most cases:o 25% of EBIT or 5% of revenue

9 http://business.financialpost.com/2013/03/19/ebay-to-overhaul-fees-in-bid-to-lure-merchants-from-amazon/

©David J. Litwiller, 2013

- Going higher usually requires more significant sharing of risk than a straightforward transactional supplier-customer relationship

- In embedded systems and consumer electronics, rates are generally lower, and sometimes much lower (0.1% to 1% of revenue)

- For technologies which are not yet commercially proven, discounts of half or more are common vs. the benchmarks above because of the risk transfer to the OEM

- For an exhaustive list of factors to consider when contemplating OEM pricing, there is a landmark IP-valuation case in the US, Georgia-Pacific, which gives a comprehensive list of factors to consider

Conjoint and Cohort Analysis

- Over time, analyze pricing by time period, by features/options selected, and by discount w.r.t volume

- Conjoint analysis assists in identifying features for which pricing can be raised, as well as ones for which customers are not willing to pay in sufficient amounts

- Time and volume discount analyses show which segments, territories and applications are being priced with better yield, and where there is likely leakage

Revenue Recognition

- Invoicing and collecting money is one thing, being able to claim it as revenue is another

- Accounting rules are complex, particularly for software with embedded service, or for multi-part delivery

Human Biases and Motivations in Pricing

- Be cautious about pricing for free, even if you discount to free – people see more value in things when there is a nominal price

- People are more motivated by getting something additional for “free” than discounting on the base item for an equivalent total exchange

- Double or compounding discounts drive greater purchase behaviour than a single large discount

©David J. Litwiller, 2013

- Always have a higher price option than the target option. This makes the target seem more economical and sensible, even if the high price option is a positioning, PR or otherwise outlandish suggestion

- Generally want to frame the target option with a higher and lower offering – the magnetic middle

- Never price based on future considerations, rather than current payments or commitment to payment. Over time, there is a strong predisposition that the giver and the getter will come to see the value diverge over time, making it harder to reach subsequent agreement.

- Premium pricing needs to be matched with discussion about uses of the additional margin: future roadmap, brand building

Price Negotiating Suggestions – Large Ticket Items, especially in B2B SaaS

- Be ready for a buyer of a large item to squeeze you on price several times. The squeezing will only stop when concessions get scarce. Plan your concessions at the outset of negotiations.

- Negotiate with a preference toward concessions on professional service hours which will give you insight and usage impact knowledge, rather than discounting the core product

- Counter profit objections with the need to build a strong support organization as well as future roadmap

- Triangulate with at least five people in the customer organization to know if you are likely winning or losing a major deal, as well as to increase the likelihood of understanding all leverage points for value and pricing strength

- If you are responding to a RFI/RFP and you were not engaged early with the decision makers in the purchasing organization to shape the elements and form of the RFI/RFP, you have a very low likelihood of winning the deal, and a price negotiation is likely proxy for being column fodder

Auction Pricing Models

- English Auctiono Raise prices as the auction proceeds o Taps into human bias for social proof, loss aversion, consistency bias, low starting point

draws in more bidders and energy, all which fuels competition

©David J. Litwiller, 2013

o Best for one seller, many buyers

- Dutch Auctiono Lower prices as the auction proceeds, o works for one seller many buyers, as well as many sellers, one buyer – o If buyer traffic is constrained, social proof behaviours are much weaker, and starting high

typically works best

- Sealed Bid Auction, forces people to guess without benefit of knowing what others are bidding – best in situations of extremely wide and subjective price variability

A Few General Thoughts to Close the Talk

- Asymmetry of resistance to price movement over time: Easier to lower than raise prices

- Simplicity: closer to consumer, the simpler pricing should be, and similarly in single person decision business purchases; however, in large enterprises and committee-based purchasing, there are situations still where one intentionally wants complex pricing or tiered pricing in order to find the optimal mix of features to capture full value and still strike a deal

- Entitlement pricing, purchasing agents – never let people see a lower price on a price list or online than what you want them to pay, since the lowest written price often forms the entitlement price from which purchasing agent bonuses paid for achieving further discounts are calculated

- You’ll never know what people want to pay until you ask them for money. Hypothetical discussions with users and surveys have comparatively little predictive value when it comes to pricing and willingness to pay

##

©David J. Litwiller, 2013