Hacker News new | past | comments | ask | show | jobs | submit login
Ask HN: Is a starting an enterprise software startup worth the hassle?
70 points by diminium on June 30, 2012 | hide | past | favorite | 20 comments
I was talking to a few people who worked for a big company. The software they use are some of the worst software I've ever seen. The workers there hate it and it makes them unproductive. Sounds like a good place to make a start-up, right?

That is until I found out the people who use the software have virtually no input in purchasing it. Not even the IT people who have to maintain it has any input in it. All purchasing decisions are made three states away in a purchasing department that buys stuff by list of features - whether or not those features work efficiently or even correctly for that matter.

How is this maze navigated? Is it worth the hassle if you don't have deep contacts and relationships inside every big company?

This is a very difficult market to break into, and lots of people are trying right now.

My first piece of advice is to build up your size and trustworthiness as much as possible. Large companies won't buy from you if you have no standing in the market. That's what you need most, after you have quality software to sell of course.

I'd recommend starting with smaller companies and building up from there—the long tail contains a huge untapped market that you can take advantage of and build up your size.

Next, know your domain exceedingly well. Don't show flashy software, solve problems. Find out what the true problems are with their current software, and solve those. Unfortunately, their problem is NOT that their users don't like using it (well, it's a problem, but not the top problem). The top problem is likely that it isn't fully capturing their process, or it misses certain savings or benefits to the bottom line that it should make obvious. Business problems are the real problems, never forget that.

Don't be afraid of investment. Use it to grow when you need to, because your small size will be your biggest downside. Even through all that, never forget that your software quality holds up the rest of your business in the end; you must balance all these factors at once.

And lastly, buy a really sharp looking suit. Sorry.

Now's actually a pretty good time to tackle these problems since "users" are becoming accustomed to using tools they like. You just have to be clever about how you tackle a particular problem, you're not going to take down PeopleSoft by trying to sell companies a PeopleSoft replacement. Rather, you're going to build something like Expensify that saves individual employees from some of the more irritating things they may have to do with PeopleSoft.

A good general trick is to build something that solves some portion of an enterprise's need, make it free to try out and cheap to use at work. There are a huge number of people out there that can pay some nominal amount for a service and expense it, it's just a matter of finding some need people with that level of autonomy have and getting them to pay you to fix it. Easy, right?

I wouldn't start by trying to navigate the maze, just avoid it altogether until you have a compelling reason to take that burden on.

> you're not going to take down PeopleSoft by trying to sell companies a PeopleSoft replacement. Rather, you're going to build something like Expensify that saves individual employees from some of the more irritating things they may have to do with PeopleSoft.

Brilliant way to put this. The way to compete with old enterprise software is first by not doing it the way they're doing it. I would start with something that appeals to individual employees, make their work easier, even joyful, and eventually you'll work up to top level decision makers.

By then you'll have a fantastic user-centric product and a list of features an exec team would drool over.

Old school: Big sales forces addressing committees and execs, operating over long sales cycles. Rewards: big contracts and institutional lock-in.

New school: Self-serve sales winning grass roots support from front line workers. Rewards: high customer loyalty, quick sales cycles, quick feedback, and lower sales costs:


I'll start off by saying I'm a closeted entrepreneur in my heart, however by day I've been selling Enterprise Software for 10yrs+. The latter is what's kept me from doing my own startup for a decade, the income is addictive but the business is not straightforward. To give some context, the average deal size for me is around $250k, but I generally work on large enterprise deals that run $2M+. Two years ago I sold over $70M in software to just two companies, neither Fortune 500. Needless to say income is substantial in this type of sales where base salaries are easily $100-$150k and you can find yourself earning at least $250k in a slow year, and as good as maybe $1M in a very good one. However, I'm different than most enterprise software sales people. Before selling I spent years programming, architecting systems and continue to work hard to keep my skills up b/c I love this stuff.

Marketing and selling software to enterprises is the most dysfunctional cycle of decision making you will ever see in business. In almost all cases the individuals in the business who actually derive value from the technology are entirely disconnected from those who make the purchasing decisions. Many times those purchasing decisions are made based on "perceived" synergies with other software systems already owned from the same vendor, having never been vetted by the actually consumer in the business. Enterprises represent a huge market to be served by well designed/functional applications. Keep in mind that most Enterprise SW contracts have customers paying yearly maintenance charges that are 20% of the original license cost and sometimes more...

The thing to understand in selling into this market is you need to adapt your "logical" marketing and sales strategy to the way enterprises buy. I say adapt b/c you can't make or help a dysfunctional buyer behave differently. These companies like centralizing technology decisions, large multi-year commitments and have complex IT accounting considerations. You need to be aware of their existing legacy infrastructures and have a story of how you "play nice". You need to have partnerships with existing large Enterprise SW companies (IBM, HP, Microsoft, Oracle,etc) and you need to have Rolex wearing sales guys who have the connections and can get deals done - http://techcrunch.com/2010/11/13/new-enterprise-customer/

If you do all that, just moderately well, a large SW company would swoop you up really fast. They can't innovate new marketable products worth crap, all they do is buy. And although we hear about some of those acquisitions there are many, many more that don't get as much press...

However in contrast to the approach above I have been whiteness to a handfull of startup business models that have done really well at infiltrating the enterprise based on their distribution and user engagement model. Take Yammer for example, their model allowed for people in large companies to start using it (for free) without having to get any approval from their IT departments. I had a CIO friend of mine who never heard of Yammer until I mentioned and showed him where over 100 of his company employees were already active on it. Large companies have to worry about compliance, document retention, etc. His first reaction was to block it via the firewall, but then he was smart enough to see it as an opportunity to invest in a technology that had actually been proven to deliver value by people within his own company. Software or services that can get into the enterprise like a cockroach, breed and spread under the radar have a really good chance of getting a CIO's attention in a good way - it's an opportunity for them to invest in a solution that works and guarantees that their business will get value since its already been proven on an small scale.

Thank you for such an amazing first comment after 966 days on Hacker News! Please keep them coming...

"The thing to understand in selling into this market is you need to adapt your "logical" marketing and sales strategy to the way enterprises buy. I say adapt b/c you can't make or help a dysfunctional buyer behave differently."

Is it easier to sell towards the good inclinations or the bad inclinations of a company?

Good inclinations is pretty obvious for us in the tech start-up scene. Think Yammer.

Bad inclinations is basically playing on fear and laziness of the large corporation. I know of a small company that makes pretty good money selling what is basically an advance employee tracking system that enforces internal security policy and procedures. It's a nice way of saying "Let's track all employees to quickly discover threats like Yammer before it spreads amongst the company to create a security hazard". Or, "we must force everyone to go through our department because that's what policy says thus using your software to do so". Of course, the reason they don't want to use this department is because they are more dysfunctional then the rest of the company. Software which enforces the bad policies and procedures of the company but allows them to "feel" more secure.

Fantastic answer.

It is not impossible but just takes its own cycle and hard as @diskdoctor has mentioned above.

To add to this there are a bunch of questions that you should answer well before Indirect Purchasing department (yes, the department that makes the purchasing decisions on all items that are to be used internally).

1. What is your business continuity plan?

2. Annual revenue. Would you remain in business long enough?

3. Technology stack - what does it take to maintain it?

4. Does it align with their tech stack recommended by architectural team? (not necessary if there is a good justification).

5. Partnerships / integrations.

6. Customer references of similar scale.


The best way for a startup could be the Yammer way - make it very attractive for users within the organization to drive it bottom up or have some real strong partnerships to drive it through.

thanks, I am glad to could contribute something here... I have often enjoyed the thoughtful dialog this community provides...

skrish, I agree, especially for a community such as Hacker News the Yammer way is the best way. If I were to start a company right now I would look to build something that an individual within an enterprise could acquire and get value from w/o involving their IT departments. As a bonus I would try to engineer some type of social, sharing or collaboration element where that individual would get exponentially more value if their coworkers also used the technology. I would closely study the technologies enterprises use at their core and figure out a way to expand, enhance or even exploit. As an example companies like Geckoboard or Leftronic could find opportunities in the enterprise by providing tight integrations into Salesforce.com, workflow/rules (Pega) data warehouses (DB2, TeraData, etc) and analytics (Cognos) - that is how large companies like IBM, PG, and GE run and manage large parts of their organization. And they could charge 10x for those features and companies would pay. CIOs like to buy software that will logically fit within their enterprise architecture. Sure, systems like DB2, SAP, Websphere, Cognos, OmegaMon, Dynatrace and Linux on z/OS (Mainframe) don't sound as sexy as working with MongoDB, Redis and Node.js but in my experience that's where the real money is. Some would argue that the future isn't DB2 or Linux running on a mainframe, but take a look at DB2 it supports in memory key/value stores. Linux on Mainframe is probably the most stable, cost effective way to run Linux on a large scale, especially if you must have a mainframe for legacy CICS apps anyway. No CIO of a large company is going to stake their career on something mission critical that relies on Cassandra or MongoDB, they go with something that works with DB2 b/c its their corporate standard and you simply don't get fired for "Buying IBM". They may experiment with new technologies on noncritical tactical projects, but face it that's not where the real money is. Companies like iRise, Roambi, Apptio, ITKO (acquired already), Splunk, Jive and Okta have the enterprise "hook" figured out in my opinion... Sure, some are "mature" now but they all started as startups...

rdl, on hiring sales people... We are a unique bread as we are hungry and heavily coin operated. The key to hiring sales people is finding those that have a track record of success. Track records not only demonstrate their ability to do the job, but they have existing networks, know how to build partnerships and most importantly know when and where not to waste their time and your time. So where do you find us/these people? Search Linkedin profiles, poach within other enterprise software companies and your network of other founders/mentors/executives who have been sold to by or employed good sales people. I was a sales manager for Rational Software for several years, the latter is how I filled almost every opening. Interviewing sales people can be tricky, they are trained to "sell" and "close". I usually look at a couple things... 1) Do they job hop? a new job for a sales person every 2-4 years is ok, but any shorter may indicate they can not sell and just keep getting pushed out. 2) Demand to see their W2 to validate past earnings. This may sound overboard but every software company I've interviewed with recently asks every sales candidate for that. The W2 doesn't lie, you want to hire the person that consistently makes $300k+ a year. They know how to do the job and they also know where not to waste their time trying to sell something they don't understand well enough.

What kind of startup would you do? Something in the "individual SaaS sales to get into enterprise accounts (e.g. Salesforce)" model?

How do you suggest startups hire those salespeople if they move into enterprise, other than hanging out at the US Open and rolex dealers?

I am a SWE by day at bigco you've heard of. I've thought about trying out enterprise software sales. It seems to me that it may be easier to dream up enterprise opportunities given more customer interaction. Would you mind if I contacted you offline to learn more about the enterprise sales side?

I find this post on the topic to be interesting: http://www.julesmaltz.com/post/3376586698/if-you-like-saas-t...

I think if you don't have the sales rolodex to sell traditional enterprise software going freemium or low cost SaaS is the way to go. Or to team up with some one who can sell into enterprises. You may be able to learn the ropes as you go but that is unlikely. If you look at the background of a lot of the top enterprise software entrepreneurs, you will see that they have worked at other enterprise software companies previously.

Like anything, it is difficult to give a sweeping answer. There are a lot of great opportunities in enterprise software that are worth pursuing. The "hassle" of selling can be reduced by have more consumerized enterprise software but selling will be a big issue. Plus it'll be more tricky as you attempt to sell to larger organizations.

A few points from my experience building a startup for the enterprise (these examples are based entirely on my experience in the financial services industry, although from what I understand they should be easy to extrapolate across other fields):

The sales cycle can be/is excruciatingly long. Certain companies that signed on took 1-2 years to get on board after initial engagement.

You need to partner up/find someone who has the deep contacts/relationships in the big cos that would be your clients.

You need to prove to them that it will be cheaper (including the cost of switching, training, etc...), Reduce Risk, and allow them to "reallocate resources."

Inertia is very strong at large organizations. Building a product that directly competes with an entrenched competitor sounds like a very risky endeavor.

I am starting an "enterprise" software company. I agree that often enterprise software is complete garbage, and the reasons organizations purchase said software is not because it is actually good, or even does what its supposed to.

But I don't think that should be a deterrent to building an enterprise software startup--there's room for disruption, and room for good software, good people, and good decisions. There are good organizations out there to work with.

Smart startups can beat the big entrenched code.

That said my co-founder is a great salesman (I'm the techinical co-founder), and having a sales team is high on my priority list.

I'm also building software for the enterprise and as the technical founder, I get excited about building a sales organization.

I'm also the founder of an enterprise focused startup[1]. We should start a club or something, looks like there are a few of us here. :-)

[1]: Fogbeam Labs, the up-and-coming Open Source collaboration / knowledge-management company that's bringing the Semantic Web + Social Web to the Enterprise. http://www.fogbeam.com Go there now and sign up for our newsletter.

Well, a new accelerator Acceleprise (acceleprise.vc) just started up to tackle the space. At the very least, it might be worthing reaching out and talking to some of their mentors.

Note: I have no affiliation with Acceleprise

My question is what industries these dysfunctional enterprise companies who buy terrible software are in and whether those industries could be disrupted.

Anything is worth the hassle if the rewards make up for it. If you don't believe me think about cameras and photo fanatics. You just have to beat someone at something to give yourself an edge. Then work that edge.

Don't attempt top-down enterprise sales unless you are ready for long, expensive cycles.

Instead, look at what Yammer has done with a bottoms-up, freemium customer acquisition approach.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact