Other People’s Money

This is the most memorable 2x2 I’ve ever come across.  It’s fundamental to understanding government behavior, consumer behavior, and business behavior. 

image

From Free to Choose by Milton Friedman

———-

I thought of this 2x2 the other day when I was discussing value propositions with a few founders.  One of the companies had an excellent product that could have saved companies tens of thousands of dollars.  The problem was that it  required a bit more work on the part of the employee at the company.   The value prop there was all wrong - when selling to big business, cost-savings isn’t the best value prop.  

When you’re selling to an enterprise, your individual customers are spending “someone else’s money.”  They don’t act like an owner, they act like an employee.  Employees want their jobs to be easier — they want  lower effort.  Sell them on that, and charge them a lot of money for it. 

When you’re selling to a consumer (or a very small business), your individual customers are spending their own money.  Low prices are absolutely essential.  They will shop around and ensure that they get the best value.  

Awesome quotes from The Effective Executive

I recently finished reading The Effective Executive by Peter Drucker.  It’s a business classic that’s chock full of exceptional advice and insights.  It was originally published in 1966 and all of the lessons still apply.  Here are my highlights.  Enjoy. 



The Effective Executive (Harperbusiness Essentials) by Peter F. Drucker



"The fewer people, the smaller, the less activity inside, the more nearly perfect is the organization in terms of its only reason for existence: the service to the environment."


"An organization, a social artifact, is very different from a biological organism. Yet it stands under the law that governs the structure and size of animals and plants: The surface goes up with the square of the radius, but the mass grows with the cube. The larger the animal becomes, the more resources have to be devoted to the mass and to the internal tasks, to circulation and information, to the nervous system, and so on."



"The danger is that executives will become contemptuous of information and stimulus that cannot be reduced to computer logic and computer language. Executives may become blind to everything that is perception (i.e., event) rather than fact (i.e., after the event). The tremendous amount of computer information may thus shut out access to reality." 



"Whether this theorem is valid or not, there is little doubt that the more people have to work together, the more time will be spent on “interacting” rather than on work and accomplishment."



"Few executives make personnel decisions of such impact. But all effective executives I have had occasion to observe have learned that they have to give several hours of continuous and uninterrupted thought to decisions on people if they hope to come up with the right answer."



"I have yet to see an executive, regardless of rank or station, who could not consign something like a quarter of the demands on his time to the wastepaper basket without anybody’s noticing their disappearance."



“What do I do that wastes your time without contributing to your effectiveness?” To ask this question, and to ask it without being afraid of the truth, is a mark of the effective executive.”


"The man who focuses on efforts and who stresses his downward authority is a subordinate no matter how exalted his title and rank. But the man who focuses on contribution and who takes responsibility for results, no matter how junior, is in the most literal sense of the phrase, “top management.” He holds himself accountable for the performance of the whole."



“If I had a son or daughter, would I be willing to have him or her work under this person?”      (i) “If yes, why?”      (ii) “If no, why?”



"They have learned that there are only three explanations for an “indispensable man”: He is actually incompetent and can only survive if carefully shielded from demands; his strength is misused to bolster a weak superior who cannot stand on his own two feet; or his strength is misused to delay tackling a serious problem if not to conceal its existence."



"He always assumes that the event that clamors for his attention is in reality a symptom. He looks for the true problem. He is not content with doctoring the symptom alone."



"One of the most obvious facts of social and political life is the longevity of the temporary. British licensing hours for taverns, for instance, French rent controls, or Washington “temporary” government buildings, all three hastily developed in World War I to last “a few months of temporary emergency” are still with us fifty years later. The effective decision-maker knows this. He too improvises, of course. But he asks himself every time, “If I had to live with this for a long time, would I be willing to?”



“A country with many laws is a country of incompetent lawyers,”



"All military services have long ago learned that the officer who has given an order goes out and sees for himself whether it has been carried out. At the least he sends one of his own aides—he never relies on what he is told by the subordinate to whom the order was given. Not that he distrusts the subordinate; he has learned from experience to distrust communications."

SaaS + Marketplace

One of the hardest things about marketplaces is that you typically only get a very small rake.  10% of services is common.  After transaction fees, you may be left with just 7% of gross services.  So, even if you’re doing 100M in gross services, you’re really still doing less than 10M in net revenue.  Some marketplace businesses have started monetizing at least in part through higher-margin software fees. 

There is a new trend emerging which combines two of the most beloved business models in tech.  

SaaS + Marketplace

SaaS (software as a service) models are very well understood and straightforward.  A customer pays a monthly fee to be able to use a hosted software solution.  

Marketplaces are also very common business models.  You simply take a percentage of all the transactions that you facilitate through your platform. 

What happens if you combine these things together? 

The SaaS gets better. 

One of the issues with selling SaaS is that the tools often enable a new use case in the company that the company is not prepared to staff.  Take Optimizely as a hypothetical example.  They provide a great tool for A/B testing, but maybe the internal resources are not in place to actually design a whole bunch of new landing pages.   This creates a hurdle for a company to adopt the software solution that may create longer sales cycles or lower sales conversion. 

Now imagine Optimizely + a Designer Marketplace.  This becomes more of a complete solution and a company can adopt both the software and the designers to start testing pages immediately.   Additionally, it ensures that they will get the full value from the great SaaS tools. This fits with Chris Dixon’s theme of owning the “full stack.”  

Not only that, all of the designers on Optimizely’s platform will be very experienced using Optimizely so they can take full advantage of all the features and functionality.  They will also likely be product evangelists and can help the classic land and expand type SaaS strategy.  Maybe even comp them for upsells of the product within organizations they’re working for. 

Summary: 

  1. Sales cycle decreases 
  2. Sales conversion increases
  3. Value from SaaS increases

The Marketplace gets better. 

The opposite happens sometimes when you start with a marketplace only.  Frequently, the marketplace may be a great source of talent, but then software is needed to manage the process and people. 

Now imagine a designer marketplace without any software solution.  There may be great designers on there, but if you hired them to do A/B testing work, you’d have to figure out an additional software package to use, train them on that tool, and get their designs incorporated into your website.  Basically, a lot of effort.  A SaaS tool like Optimizely makes the company and the designer more productive.  

The key to having the SaaS and the Marketplace be additive to each other is that the software must increase the efficiency of the work on the marketplace.  People need to prefer working with your tools vs working with other tools.  It needs to benefit both the buyers and sellers on the platform.  

The SaaS tool may also be an effective lead gen source to the marketplace.  If you can make it simple to get design work done through the platform, it will dramatically increase usage of designers.   In those cases, it may make sense to offer the software for free to get users into the marketplace.  It depends how you want to monetize the business.  

Summary: 

  1. Increased productivity
  2. Increased volume

———-

I believe the trend towards SaaS + Marketplace will continue.  

Up until today — many software companies launched programs to partner with “solutions partners” - typically small consulting shops.  The problem with this is that it places the burden of selecting and managing providers on the customer rather than the software company.  This is the way software has been sold in the past, but it will change as customers expect more out of their software vendors.  Customers expect great software and great integrated services.  

What type of marketplace are you?

A founder proudly told me yesterday that “Liquidity is not a problem.  71 people applied to a job within 24 hours!”

I replied, “That sucks… 70 out of 71 of your users had a bad experience.” 

—————-

This led me to think about the different types of possible workflows for marketplaces.  I wanted to share my thoughts on a framework for thinking about the different types of marketplace matching mechanisms.  It’s absolutely critical to think through the correct workflow for matching in a marketplace.  There is no single right answer - it all depends on the service that you provide. My experience has also been that almost everyone starts with an overly complex workflow.  The simpler and faster you can make the matching process, the faster you’ll be able to grow.  

Double Commit Marketplace

Examples: Care.com, oDesk, Thumbtack. 

A typical workflow:  Buyer posts job.  Buyer invites candidates AND Candidates apply to job.  Interviews happen.  Buyer makes a hire. 

These marketplaces are the easiest in some sense because all of the curation, filtering, and matching is really done by the users.  They also allow for completely custom needs to be fulfilled since people can chat and interview and be flexible.  On the other hand, these marketplaces tend to have the lowest overall fill ratio (hires made / jobs posted) since there is substantial time and effort required by both sides.  They are also the least efficient overall since much time and energy is “wasted” on applications and interviews that do not turn into real work.   As a result, one of the big risks is that you end up with a vicious cycle of quality.  If high quality suppliers get frustrated with the process, they will leave.  

The key focus for the company in this case is often on improving the fill ratio.  Lots of emphasis is often placed on streamlining the candidate -> interview -> hire process since there is significant dropoff through this funnel.  

————- 

The next evolution of marketplaces is to single-commit models, where one side can make a transaction without the active approval of the other side.  There are two types of single-commit models — you could have the buyer pick or the supplier pick.  As soon as a marketplace moves to a “single commit” model, they must be responsible for the quality of at least one side of the marketplace.

Buyer-Picks Marketplace

Examples: YourMechanic, Unwind.me, Airbnb. 

A typical workflow:  Suppliers enter availability.  Buyer can see available suppliers.  Hires a supplier without discussion.

These marketplaces are a bit more complicated since additional availability criteria must be captured from the suppliers.  It does not fit well with low utilization marketplaces.  For example, if a mechanic was only getting 5% of his work through YourMechanic — he is unlikely to keep his schedule up to date and be responsive and available when anyone wanted to book an appointment.  In addition, the suppliers in a buyer-picks marketplace can often get frustrated because they have no control over their own utilization rates.   It’s important for a buyer-picks marketplace to keep a close eye on utilization rates

I include Airbnb as a buyer-picks marketplace since they are increasingly trying to move to Instant Bookings for properties.  These consist of a curated listing with professional photos and an up-to-date calendar of availability.  It also requires that Airbnb becomes at least somewhat responsible for the quality of the buyers.  If they send offensive, dirty, rude, etc buyers, then the suppliers will leave the platform or at least no longer allow instant, unapproved bookings.  

The key emphasis here is typically on quantity and quality of listings and maintaining accurate availability of suppliers so that buyer demand is always filled. 

Supplier-Picks Marketplace

Examples: Uber, Rev.

A typical workflow: Buyer posts job.  Approved suppliers see available jobs.  Supplier claims job. 

These marketplaces require the highest degree of job standardization and quality control by the marketplace.  Since a high degree of standardization is required, these ones must go after a specific vertical, such as usability testing. With respect to quality, in an open double-commit marketplace, poor quality results are not blamed on the company.  However, in a supplier-picks version, all of the blame will be put on the company if there   is a negative experience. On the plus side, these marketplaces can have by far the highest fill ratios since it is possible to drive towards 100% of jobs getting claimed by suppliers.  

The key emphasis here is on maintaining an active and curated pool of suppliers so all jobs are done quickly and effectively. 

The Pros and Cons

image

——————-

I think whenever possible, you should try to be a supplier-picks marketplace.  They have the highest potential growth rates and can have the best overall user experience (highly correlated with low effort and high quality).  They are also the hardest though, they require constant attention to balancing supply and demand and running exceptional operations.    

Liquidity Hacking - How to Build a Two-Sided Marketplace

This post was originally made as a guest post on PandoDaily and VentureBeat.

Marketplace businesses… They always seem great on paper, but it’s so insanely hard to solve the chicken or the egg problem. Every founder I meet who’s building a marketplace business basically says the same thing: “It’s so much harder than I thought to build a two-sided marketplace.

But of course it can be done. There are plenty of examples of success. EBay did it. We did it at oDesk. OpenTable is famous for doing it, and AirBnB is killing it. What I’ve noticed in the success stories is the basic theme that I call “liquidity hacking.”

What is liquidity hacking? 

In short, liquidity hacking is the strategy that successful marketplace businesses employ to reduce the challenges associated with low transaction volumes. It often requires narrowing the scope drastically of the offering until sufficient scale allows you to expand to achieve a broader marketplace.

This type of hacking is completely necessary in my opinion. The thing with marketplace businesses is that they work great at scale. Once you have hundreds of thousands of users on the supply side and the demand side, then everything works great. The challenge is almost always — how do you get there?

Provide value to one side

Offer portfolios — Many different people like to showcase their work online. In the olden days this would be done with a custom and personalized website. Now there are free portfolio tools for almost every type of individual. Once a site is successful in gaining users for their free portfolio tool, you end up having solved your initial supply challenge. See: BehanceDribbbleCarbonmade

Build community — If you can establish a healthy and vibrant community on either side of the equation, you’ll end up with the potential for a great marketplace. See: StackOverflow
Offer Tools — If you can get a lot of people to use your free tools, then you’ll have a highly engaged user-base on one side of the equation. See: GithubOpenTable.

Find aggregators

Find physical aggregators — There are various aggregators in the real world that provide really novel ways to hack liquidity. These are physical locations that contain a concentration of supply or demand. College campuses work great for this. Word spreads lightning fast and student labor is crazy cheap. I’ve seen some marketplace businesses that start focused on campuses and figure out the formula for getting significant penetration. Then just go to the next campus and repeat. Basically same deal with large office buildings and even high-rise apartment buildings. The great thing about large office buildings is you can sometimes convince the HR manager that your service is so cool that they should email the entire company about the offering. Now you have a trusted recommendation to try a service and you get free distribution to all employees. See: MyEatClubZimride

Find an enterprise client — Enterprise clients often have lots of money to spend on the demand side. Sweet. A single contract can get you off the ground and running and focusing all your efforts on building the supply side and more product. The risk of this is that the enterprise feels like they own you and demand increasingly complex product features that could ruin your platform for other buyers. If the relationships are managed well, it’s a good way to get started. See: Gigwalk.

Find supply aggregators — Let’s say you need a whole bunch of freelance developers. You have a vision of cutting out all middlemen and creating real value in the world. But you know it’s going to take a long time. So, you look for some supply aggregators like a Web design firm in India to offer the supply you need to get off the ground. Or maybe you have a vision that anyone should be able to rent anything from anyone else. That’s great, but you need a lot of rental supply to get things started.  What better source than places like Home Depot for power tools and REI for camping gear to get you off the ground with a solid supply. Then you can spend your time worrying about demand and about getting supply from individuals or owning your own supply later. See: Getable

Scrape listings — For example, scrape a whole bunch of real estate listings from Craigslist. Then put up a modified version of that listing on your website with a better UX. The problem is you can’t actually connect the supply and demand since you don’t own the supply yet.  But you’ll get damn good conversion rates if you say, “Hey supplier, good news. I have a customer for you, you just need to sign up for free here.”

Narrow the Problem 

In almost all cases, it helps to really focus the efforts of the business. You have very limited resources to get to a sustainable marketplace so you should concentrate your resources on small areas.

Focus on a geography — This is kind of obvious. See: every local marketplace ever.

Focus on a niche community — Beanie baby collectors. See: Ebay.

Focus on a vertical — You started with anything? Now making a standardized offering around housecleaning. See: Exec.

Curate one side

One way to make things function reasonably well at low volumes is to curate one side.  There is a question of whether a “curated marketplace” is really a marketplace at all, but so what? Curation is really important. The reason it’s so important is that it greases the wheels of the marketplace. It allows the marketplace to move from a listings model to an open-call transactional model. This brings a lot of efficiency to the process and lowers the effort required by the buyers. YourMechanic is a great example of this — rather than just have a directory of mechanics that you have to review, they presented me with just two options and their calendar of availability. I didn’t have to think about who to hire, I just clicked on the first available slot since they both looked like great mechanics. See: YourMechanic

Use “hamsters”

When all else fails, use hamsters. This is basically what I was for my first year at oDesk. It’s the brute force method. You just throw bodies at the problem until you have enough scale that you can reduce the manual requirements in the process. At oDesk we always had the vision to be kind of what you see today, but in the beginning we acted more like a staffing firm, and I was the recruiter that would talk to the customers to take the job requirements and then go on a manual search to find candidates to fill the position. We just had to hustle like crazy, because you only have ~24 hours before the demand side loses interest and realizes that you have nothing. See: oDeskAirBnb.

Note: If you just want to know which one came first, chicken or the egg… It’s the egg. From Wikipedia:

The theory of evolution states that species change over time via mutation and sexual reproduction. Since DNA (deoxyribonucleic acid) can be modified only before birth, a mutation must have taken place at conception or within an egg such that an animal similar to a chicken, but not a chicken, laid the first chicken eggs. These eggs then hatched into chickens that inbred to produce a living population.

What On-Demand Teaches All of Us

Liz Gannes published a well-done series on the on-demand economy on Recode.  She talks about the rise of food delivery startups, on-demand rides, on-demand laundry services, etc.  There is certainly a huge wave of these startups that are all harnessing the power of having every worker walking around with a GPS-enabled smartphone (as I’ve said before, the smartphone is a great logistics device).  

Many of them are taking off and making super-happy customers in the process.  There’s something magical about the first time you order from Sprig and your dinner gets to your door in 10 minutes flat compared to the typical 45-90 minutes that traditional restaurants take. 

The on-demand economy has taught us all that turnaround time is always important.  Even if your service normally takes 1 month to deliver, if you can do it in 2 weeks, you will amaze customers.  It doesn’t have to be “on-demand,” but you should always strive to improve the speed of your service.  Nobody will ever want it to be slower. 

Sometimes it’s very difficult to deliver massive cost savings or massive quality improvements, but turnaround time improvements are frequently achievable.  Analyze everything about your process and find ways to cut wasted time. 

Speed should be a key performance metric of every business service, from taxis to analytics software. 

Speed applies to everything.  We always want things to be faster.  It applies to business software solutions just as much as it does to consumer services.  

A few examples: 

———-

Omniata delivers analytics and engagement software.  The performance is incredible - processing billions of events and allowing custom reports in fractions of a second.  Just the simple notion of being able to create reports and see instant results has caused a behavior change in users. They are more likely to create and use more reports, and therefore get more value out of the platform. 

———-

At oDesk, we found improvements in customer satisfaction when we decreased the time between a job posting and the first supplier application.  In the beginning, everything was manual and it would frequently take us 24-48 hours and a few phone calls to get the first applicants to a job.  Once we built and scaled the marketplace, we were able to get those first applications in a matter of a few minutes which was amazing for customers. 

———-

Google took their blazing fast search results and just kept trying to make them faster.  Here is Marissa Mayer on her usage of Google Instant

“One of the things I’ve seen in my own personal usage,  is that while each search is faster, I spend more time doing searches. Because I actually see the results coming in and out as I’m doing my searches… I learn things as I go. And after I’ve actually fulfilled my query, a lot of times I’ll see interesting suggestions, so I’ll scroll around and learn different things and so I think ultimately, it may increase engagement of our users.”

-Marissa Mayer

———-

Jeff Bezos has known it all along too. 

"What’s not going to change in the next 10 years?’ [I]n our retail business, we know that customers want low prices, and I know that’s going to be true 10 years from now. They want fast delivery; they want vast selection.  It’s impossible to imagine a future 10 years from now where a customer comes up and says, ‘Jeff I love Amazon; I just wish the prices were a little higher,’ [or] ‘I love Amazon; I just wish you’d deliver a little more slowly.’ Impossible. And so the effort we put into those things, spinning those things up, we know the energy we put into it today will still be paying off dividends for our customers 10 years from now. When you have something that you know is true, even over the long term, you can afford to put a lot of energy into it."

-Jeff Bezos (full quote and more here)

———-

Davy Kestens at Sparkcentral has known that speed was critical since he started his first business.  He personally provided customer service solely through Twitter.  He replied in a matter of minutes to customer support requests instead of typical customer service centers that offer “ticket response times” of under 24 hours.  24 hours?!?!  That seems like an eternity nowadays.  Sparkcentral helps large customer service organizations respond to requests in a matter of minutes, providing a huge speed improvement for end customers.  

———-

So, what is your speed metric?  How can you constantly increase the speed of your service? 

Why I’m Short $CRCM

How’s this for a checkout page: 

image

This is terrible and distracting design, but it’s worse than just that. 

My biggest issue with Care.com is one of misaligned incentives.

Care.com makes the bulk of their money from paid memberships of customers and caregivers.  The average lifetime of a paid subscriber is 7 months according to their S-1 filing.  Also less than 10% of their users overall pay for a premium membership, but you can’t even contact anyone on the platform unless you pay!  Think about that for a second.  That means maybe 90% of job postings on the site result in ~15 candidates per job and not a single contact or interview is actually performed.  I’d guess maybe 2-3% of all applications ever turn into an interview.  Think about how bad that experience is for caregivers - you might apply to 50 jobs just to get one interview.  They are using the free job postings and the huge numbers of futile candidate applications as an attempt to sell families on a paid subscription.  

Why not just focus on making it easy for me to connect, hire, and pay highly qualified and reliable caregivers.  If they added value after the match, they would not be so focused on trying to make me pay to talk with people. 

Take a closer look at some of the numbers below from their S-1 filing. image

(in thousands)

What should be alarming are the trends here: 

Percentage of families paying a membership is on a bad trend.  14.8% in 2010, 14.1% in 2011, and 12.3% in 2012.  

Percentage of caregivers paying a membership is on an upward trend. 1.2% —> 1.5% —> 2.0%.  However, I view this as a bad trend as well. 

The reason that I don’t want caregivers to pay is misaligned incentives. 

Care.com is incentivized to get as many caregivers to pay as possible.  It looks to be a growing trend.  However, the reason for caregivers to upgrade to a paid account is to get premium placement in search.  I hate this model — it’s featuring people that are willing to pay rather than people that will be the best caregiver.  

Care.com should put the best available people at the top of my list.  Instead, they accept bribes.

Why I love the Care Space

Startups are attacking just about every vertical.  However, I’ve been surprised to see one industry where there hasn’t been much startup activity.  Perhaps because it’s unsexy and support calls may be life or death situations.  It is an $84B industry in the US. 
The Elder Care Industry
An $84B industry represents an estimated 3M+ individual workers in the industry.  People may be hesitant to enter this market because of difficulties working with Medicare and Medicaid.  I can only imagine how difficult it is to work with them.  But there is plenty of reason for optimism.  Estimates are that there are nearly 1M workers that fall in the “grey market” - the term used here for home care aides paid out of pocket through private arrangements.  
Just for reference, there are 233,000 taxi and limo drivers in the US. ;) 
image

And let’s take a look at how much the workers earn. 
image
So, why do I like this industry so much.
Massive industry fragmentation.
As of 2010, there were 82,239 different home care agencies.  Tons of franchises exist to help people start a home-care business with little money down, usually a $75K franchise fee to get things going.  Lots more individual providers advertise their services on Care.com, Craigslist or any other local services directory.  It’s incredibly hard and time-consuming for a customer to go through all these options and find a qualified provider. 
Unsatisfied customers
I’ve talked to many people who have gone through the process of finding home care for an elder parent or relative.  Every story I’ve heard is similar.  They interviewed many dozens of providers and had a very hard time finding reliable and high-quality caregivers.   When they did find someone they liked, they frequently lost the person later because caregiver retention is incredibly low. 
I don’t have any hard data on customer satisfaction in the industry, but I’d be willing to bet that the NPS scores are terrible.  
Unsatisfied workers
For elder caregivers that are working at an agency — the main benefit is client acquisition and stability of work.  They are all working independently anyway, so they’re not getting much else.   For that privilege, the agency takes about 50%.  If you assume for a moment that the typical customer will end up using the service part-time (10 hours per week) for 6 months, the profit from that relationship for the agency is $2,400 — 24 weeks * 10 hours per week * $20 hourly rate * 50% agency fee = $2,400.  That should be way more than enough to cover acquisition costs.  
I believe that a mobile-first service leveraging good logistics and workflow could drive down the overall cost while at the same time paying the best caregivers more money and offering higher utilization.  More money and faster pay to caregivers = happy caregivers = happy customers. 

Highly recurring needs
Most in-home care is highly recurring.  It needs to happen at least every week and sometimes every day.  This puts it in one of those rare categories like taxis or food delivery that is needed very frequently.  
Baby boomers
In case you haven’t been paying attention.  America is aging.  The market for in-home care is expected to grow rapidly for the next few decades as the baby boomers start needing care services. 
image

Some big challenges
One problem is the “monogamous relationship” aspect of elder care and the relatively high degree of trust required.  This may make disintermediation a big challenge.  I suspect that this is an issue, but there are a few things that mitigate these risks.
1) Value-add from the mobile app.  The convenience factor that comes from the parent being able to book, schedule, and pay through the system may be enough to keep all of the work on the platform.  A geofence around the elders home could give simple push notifications to alert the adult that the caregiver has arrived or has left.  
2) Ability for on-demand / flexible workers.  While it may be natural to have a monogamous relationship with a primary caregiver, there are surely times when the schedule doesn’t work out and a substitute is needed.  Making this process easy and convenient should be enough to prevent disintermediation.
3) Value in guaranteed work / payments.  The brilliance of the oDesk hourly model rested on the guarantee of work.  For those that don’t know, the work diary feature of oDesk takes screenshots every so often, therefore guaranteeing that work was being performed and billing is accurate.  The same opportunity exists in elder care.  
4) Value add for trust.  A marketplace should be able to optimize for the trust and quality of caregivers far better than the individual family reviewing their options.  Just for reference, a current search on Care.com yields 1,800+ caregivers just in LA.  That’s not good at all.  How am I supposed to choose out of those 1,800?  I also believe a marketplace can build a brand around trust and quality by being very proactive about curation and adding standard services like certifications, background checks, and insurance. 
If you’re working on something in the space, I’d love to hear from you.
————
PS - In a future post I’ll explain why I’m bearish on Care.com. 

The Delusion of Customer Testimonials

I’m in the fortunate position of being able to see a lot of pitches from a lot of different startups.  One of the things these pitches have in common is testimonials from lots of happy customers.  It’s not very surprising given the point of the pitch deck.  

But, I believe the testimonials can be leading to a larger form of delusion in startups.

While helpful for the pitch deck, it is important for entrepreneurs and investors to avoid falling victim to testimonial delusion.  Testimonial delusion is when we extrapolate a few good comments to mean all is good.  As investors, we run the risk of being swayed by a false sense of social proof.  For founders, the overwhelmingly positive testimonials could lead to a false illusion of market fit.  

Here are the issues as I see them for typical startups: 

Conversion Selection Bias

A lot of people come to your site and never convert.  The fact that someone converted already means they’re positively inclined towards your product.  Something about your site and messaging really resonated with the visitor.  It’s the other 90%+ of visitors that didn’t convert that you need to know about.  

User Selection Bias

If you have 100 users and need 3 testimonials, who are you going to ask? Most founders ask their most engaged users.  SImilarly, if you were to send out a survey to all of your users, you’ll get sample bias as the most engaged and happy customers will happily spend time to fill in a survey but your “passive” (see NPS) customers will likely just ignore the survey.  This further leads you to be talking with only the most enthusiastic fans. 

Misaligned Incentives

The user submitting a testimonial has some incentive to say very positive things and hardly any incentive not to.  It’s all upside for the testimonial giver. One popular SEO tactic is to give testimonials to every vendor you work with in order to get a backlink and go up the search rankings.  Additionally, it just feels good emotionally to compliment someone and in return receive a high degree of gratitude. 

Rating Bias

I’ve talked about this before with 5-star feedback systems.  Basically, there is some simple rating bias that humans inherently have — we all tend to be nice people.  You’ll see that the vast majority of feedback systems do not operate on a bell curve.  They are A-centered or 5-star centered.  This is in part because humans are terrible at assessing the absolute value of things.  If you ask people to rate things on a scale of 1-5, you end up with lots of 4s and 5s.  If you ask people, “Is A better than B?” you’ll get a much clearer picture.

—————————

So, I believe testimonials are great and are sometimes useful, but I believe we all need to exercise caution when using them to make any decisions. 

A few things we can all do a little better. 

Focus on the Quantitative

It’s very tempting to use anecdotal testimonials and punchy quotations.  A few of those 140-character ones fit perfectly on slides. However, we should all really be looking more at quantitative metrics and objective behavioral criteria.  I see people report sometimes on surveys they ask customers, “How much would you be willing to pay for xxx?”  We all know that the reality of what people pay and what they say they will pay are two very different numbers. 

Random Customer Sampling 

When evaluating the real picture, first start by talking to a random sampling of customers.  Get a list of customers and choose at random to call the user directly and get their input.  

Random Prospect Sampling 

To address the customer conversion bias, we can get a list of prospects and try to contact them just to get input and feedback on the product.  Can always do this through things like UserTesting, Mechanical Turk, or FeedbackArmy as well in a quick and dirty fashion.  Or use SurveyMonkey’s Audience feature. Make sure to run some baseline tests on competitive websites or other similar services so you can compare your site to the competition.  It’s so cheap and easy to do. 

Allow for Anonymity

People are much more willing to be honest (and harshly critical) when they have anonymity.  Just look at Secret vs Facebook. Give users the chance to be anonymous with you - see if you get different input. 

Survey Completion Percentage 

A focus on getting the highest participation rates in surveys will help eliminate sampling bias.  I’m personally a big fan of SMS-based single question surveys.  The completion rate for these can be north of 80%.  I think this is one of the most underutilized tactics out there. 

Use Standard Survey Design 

I’m a believer in Net Promoter Score.  It’s easy to compare and benchmark your company against the industry.  On the other hand, I still see tons of companies using custom surveys and questions and varying collection methods.  The more you standardize your survey, the more useful it will be to you to benchmark against your industry.  Check out Delighted App for an easy way to launch your own. 

Any other good ideas? Let’s hear ‘em. 

The Ingredients for a Successful Marketplace

Lots of entrepreneurs aim to build a marketplace.  It’s a popular type of startup.  People look at the success of an oDesk or and Uber and think, “I’m going to build the Uber of _______” and they fill in the blank with any category from construction rentals to lactation consultants.  But, not all marketplaces will work.  You need certain ingredients in an industry to have the potential for a successful marketplace.  Bill Gurley did an exceptional job mapping out 10 key factors for marketplaces.  I have my own framework for thinking about different verticals. The following is based on a conversation I had with Semil Shah.  It’s my take on 5 key ingredients.  I left out market size discussion, since I believe you can build a successful marketplace without  shooting for venture size returns.  

Ingredient #1 - Recurring Usage

Much easier to acquire a repeat customer than to acquire a customer for the first time. 

Everyone knows that there is a huge drop-off throughout the customer on-boarding funnel.  Maybe 90% of visitors never sign up, and then a further 60% of those signups may never actually make a transaction.  So, when you do finally convert a visitor to an active participant in a marketplace, you better do everything you can to keep them engaged and making further transactions.  Of course, this is nearly impossible if you’re building a marketplace for wedding photographers since people (theoretically) only get married once.  Therefore, ingredient #1 is Recurring Usage. It is fundamental to a marketplace.  The best marketplaces have high-frequency and highly recurring transactions.  Example: Uber - most of us city-dwellers use taxis all the time, maybe several times every single week. If you’re considering a marketplace, look at the standard buying behavior in that industry: how often do people buy? 

Ingredient #2 - Irregular Usage

This ingredient is less obvious.  Credit goes to Vidur Apparao for articulating this ingredient’s key role in a marketplace.  Irregular usage is important because it: 

1) Increases the need for on-demand resources, and

2) Reduces the likelihood of getting a dedicated resource.

If you don’t know in advance when you’re going to need resources, it’s hard to plan and people would like an instant, on-demand service instead.  Additionally, if you don’t know when the need will occur, you cannot hire a dedicated resource to be available whenever you need it.  

For example, my buying habits for massages are highly irregular.  I get massages from time to time, but it is usually after some sort of strenuous workout or when my body just happens to be sore.  (Of course that seems to be happening more frequently as I get older).  Therefore, I rarely plan appointments for massages weeks in advance.  I strongly prefer to get on-demand resources within a couple hours of when I feel the need. Additionally, since my usage is irregular, I do not have a single go-to therapist that I can expect will be available when and where I happen to want a massage. 

Ingredient #3 - Standardized Work

Standardizing the work means you can guarantee the quality. 

Horizontal markets can thrive in custom environments, eg Craigslist and eBay.  But for a vertical marketplace to work well, the work or items must be standardized.  The more standardized the work, the more a company can control the quality and reduce the effort required by the customer.  If custom work is required, the buyer needs to spec out the work, select an appropriate provider, and carefully manage the process to make sure the requirements are met.  The onus falls on the buyer to manage quality. There is no way that the middleman can guarantee quality since it is subjective for any custom work requirements. In contrast,  if the work is standardized, you can typically measure the quality of the output. If you can measure it, you can improve it. 

As an example, look at Rev — they provide audio transcription and translation.    The inputs and outputs of an audio transcription are consistent - audio files go in, transcriptions come out, and we can score the results.  Rev can manage the entire process since they know what needs to be done at each step and they know the best available transcriptionist to do the job.  The result is a dead simple process for buyers that will have continuously improving quality. 

Ingredient #4 - Little Trust Required

Trust is a barrier to transacting online.  Every time you enter a transaction with a new individual or company, there is some part of everyone that worries about  getting screwed.  If the dollar amounts are small, people worry less.  If they are large, people worry more.  If the transaction involves additional potential loss, such as having your house robbed by giving away a key, the stakes are higher.  If your kids safety is involved, the stakes are at an all-time high.  This last one is why an Uber for babysitters is going to be exceptionally difficult.  Marketplaces work best when there are relatively low stakes involved in the transaction, it gives people the willingness to transact online with an unknown entity.  

There is good news on this front though… Facebook.  Facebook (and to a lesser extent LinkedIn and Twitter), do an incredibly good job at instilling trust in each other.  Airbnb has been masterful at using Facebook connect to instill trust in renters and owners.  I know as an Airbnb user, I feel 10x more comfortable renting to someone who has a well established Facebook presence and is a 2nd degree connection.  These networks increase the amount of trust we will place in an unknown entity, thus enabling more marketplaces to flourish. 

Ingredient #5 - Non-monogamous Relationships

This ingredient is a result of little required trust and highly standardized work.  The result is that the buyer doesn’t care if they have the same provider every time.  This is true for movers, for example.  It is not the case for housecleaners — and I don’t believe the marketplace for housecleaners is a good one even though it is highly recurring.  It also makes a marketplace for something like dog walkers a bit problematic.  My wife and I have a lovely dog named Isabella and we like using the same dog walker every time we need the service.  This is because we’ve given him a key to our house and he takes good care of our dog — we’re not willing to change up service provider every time we need a dog walker.  This is less of an issue for dog-sitters though, because I don’t need to handover a key to my house.

These 5 ingredients aren’t everything that you need, but they’ve always been a helpful framework for me to evaluate whether a marketplace should exist in each vertical.  

———————

So, I wanted to run a little analysis.  I spent a few hours going through Craigslist service categories and thought through 30+ based on the 5 ingredients above.  I didn’t take market size into account.

A few interesting results as I went through the categories. 

Some of the best: taxi drivers (Uber) and dog sitters (DogVacay).  Some where I haven’t seen success yet: massage therapists and travel agents. 

Some of the worst: wedding planners, interior decorators, and landscapers.