May be some of these hints are useful for some of you. Pulled together as one set you may find them here. Exactly this is what Memonic is all about: Collect snippets of information, pull them together and share them with others. Try it out yourself!

Here a brief summary of the topics covered:

1. Start with a simple plan: A (business) plan summary requiring more than one (1) A4 page is probably too complex to understand.
2. A few commonly shared rules: A set of collective values is as important for success as your product
3. Get a businessperson into your team (and vice versa): Techies don’t sell…
4. Get a professional bookkeeper: (financial) numbers matter.
5. Time tracking pays: You need to know for what tasks need what time.
6. Start with a good founding contract: If divorce looms, you’ll not fix terms and conditions à l’amicable; better fix them at start.
7. The best 10$  you’ll ever spend: All the tools you ever need to start.
8. Use cloud computing: Don’t buy infrastructure, rent it on pay-as-you go.
9. Never buy hardware at list prices – Never: To pull someone’s heart strings helps – You know, we’re a poor cash strapped startup.
10. It’s a virtual world: Virtualization of computers saves you even more cost.
11. Bend the rules: No cheating, but drive a hard bargain and watch for offers.
12. Forget tapes, do online backups: “Yesterday all my backups seemed so far away,” is no backup strategy for valuable customer data.
13. Don’t waste time that others do better:  Use those online apps out there for mail, wiki, task tracking and more (plus it’s free or almost free).
14. Get early agreement – do mockups: Nothing is worse than this comment on launch day: “Oh, this isn’t what we wanted.”
15. Do usability testing: Your team is clever, the crowd is smarter.
16. Consider using agile development methods (such as scrum): Because six month release cycles or a no-go for any startup nowadays.
17. Use puppet: No we don’t talk about a doll, but about automated server setup and configuration.
18. Automated testing: You have to break an egg to make an omelette.  So better fix your errors early one. Automated error recognition helps.
19. Catch those warnings – Use Splunk: Your own log files are true gold mines. So mine them – Splunk is your shovel.
20. 80% of success is showing up. The Olympic motto pays for startups, too
21. Learn to live with the ups and downs of every startup: Each high is followed by a low, followed by a high
22. Use memonic.com – There is no simpler way to keep the essential snippets of information

During the last weeks Memonic received a number of awards, closed a financing round, got a lot of new users, all in all a very successful period with positive energies flowing freely. A few weeks before it was all different: A few rejects at startup competitions, no-I-not-in from potential investors, user feedback they like the competition more, a negative commotion.

As a start-up team it is important to be able to deal with both moods. After a lean time follows with hard work and a bit of luck a winning streak. And even the boldest winning streak will end some time.

As a founding team it is important to celebrate success yet not loose contact with reality and deal with setbacks without giving in and giving up.

How to advertise your application? A national ad campaign is probably over your budget, so is the TV. Yet there many alternative and very cost effective methods.

One of the best possibilities to promote your product (and yourself) are the manifold events taking place. Just to name a few here in Zurich an vicinity: StartupCamp, Barcamp, WebMonday, WebTuesday, Chuchis, Venturelab Events, etc. etc. The Olympic slogan pays: Participation is everything.

As soon as your website is live a ton of data is silently gatherd in countless log files across your application’s systems. Most often nobody ever pays those log files tribute, i.e. reads them. Normally the sys admin will delete the older files once disk space becomes scarce. We think this is a bad approach.

A smarter method is to use tools such as Splunk. Splunk at its core is nothing more than a full text search index built from multiple and disparate log files. You can search for specific patterns, and Splunk will create on event an automated notification. You also can search the log files for error codes. And best: This works real time, too. A coder can watch what happens at creation of that server exception.

I haven’t seen a more clever usage of log files.

Wit the launch days away most startups focus on getting it done. If achieving this goal requires a compromise quality is often the accidental victim.

Goal achieved, successful launch past you, the ensuing spaghetti code may accompany you for a while. And as always, errors strike in the most unexpected and most unwanted moments. A simple yet powerful remedy is automated software testing.

Tools such as Selenium help you to set these tests up, run the tests and control its’ outcomes.

No, this blog post is not about a child’s doll. They’re lovely and my daughter loves them very much. This post is about a software tool – better a framework – with the name Puppet.

After months of development you application is ready for deployment. Quickly install the application on the live servers and then off to the release party. Quite often the next day brings a hangover – not because of the party.

Software development and software updates are often underestimated in their complexity. Starting with the third live server things get more complicated still.

How to deal with this? Use Puppet!

Puppet is a software framework that enables automated deployment scripts and helps you control their execution. We employ Puppet to deploy entire batteries of servers and the applications than atop.

You may be need a little help to get into Puppet. But once up and running it’s well worth it.

The traditional waterfall methods for project management have much for them: Tested for centuries. That’s how big engineering feats have been delivered, or so they make us believe.

The other day the 40th birthday of the ill-fated Apollo 13 mission was commemorated.The mission nearly ended in catastrophe after an explosion cut oxygen supply. Up there and down at mission control in Houston they had to come up with solutions fast. Improvisation was l’ordre du jour.

Both methods have something for them. For any startup it’s probably okay to apply a bit of both. You don’t want to get bogged down with development cycles that last months, neither with the improvisation à la Apollo 13.

In our case we employ Scrum as an agile development method. Other such methods are around. The basic thing it does: It forces you to be honest in short cycles (two weeks in our case). After each two week segment we know where we stand.

All your mockups (previous post) will not protect you from completely and utterly misrepresent what users really want the application to behave.

There is only one way to get to the point here: Do usability tests. You will have a very bad day that day. Your testers will not find that button you specially designed and argued so much about, they will be utterly lost in the application although you thought that the information architecture is so simple an clean.

Yet that is the way an application is perceived. Better listen early and often and adapt your approach.

Even if it requires a dose of masochism Usability Tests help!

And off you go. All set up, your product idea clearly spelled out (at least that’s what the founding team thinks), now it’s just a question of execution. Or so you think.

Countless number of times we came across situations where even seasoned teams thought they had spelled it out, only to discover after implementation that there is a notable gap between what different stakeholders thought the product should do and the implemented product actually does.

We suggest a different approach: Do mockups early on.

Put on paper (digital paper is also okay) what you want the application to look like and to work like. You will see that often the initial agreement disintegrates once you start doing mockups only to emerge with a stronger consensus on what the application really is supposed to do.

13 – Don’t waste time that others do better
(plus it’s free or almost free)

Any company needs a number of basic processes and applications such as Email, calendaring, time tracking, billing, version control, task tracking, etc. And quite a number of startups had in the past first to invest a couple of person weeks in installing your email server and all the rest (We know what we talk off… ;-(

Today that’s a different matter: All these basic applications are yours for free are almost for free. Examples include Google Apps including Email, Calendaring, Office Applications, Paymo and Mite for time tracking, Beanstalk and others for version control, Highrise for CRM, Memonic for online research.

Focus on what really counts – your product or service – and don’t waste time on what others do better.