1. Before anything start with version control and how to use it effectively. With out this step you got nothing.

2. Collaboration. Choose a project management tool (Jira or the like) and show them how to use it effectively. Show them how developers use it to keep PHB & managers honest.

3. Continuous Integration. Choose a CI of your choice, set it up, get it working for them, get them addicted to the time/work savings. (I use Jenkins, but now there are better ones)

4. Testing. Teach them why testing is good, help them implementing the testing framework (On the CI) and help institute it as a 1st level part of the development process.

My style, is to never give them "more work", but to show them how each step saves them work and time.

Regards

I don't have a reference for start-to-finish automated testing, but I like the way in which Justin Searles explains the difficulty of such a thing here: http://blog.testdouble.com/posts/2014-01-25-the-failures-of-...

I gave a brief presentation about testing a few months ago and felt like it didn't really connect. I think it's genuinely hard to break the preconceptions about testing, and also there's a fair amount of incidental framework and library minutia that can be hard to isolate from the big picture ideas.

Subjective experience is a huge factor. Trying to rationally explain the advantages or conceptual underpinnings of testing doesn't always work. (In hindsight it didn't work for me either.) It really takes sitting down next to somebody and walking them through a process. I've had success in showing a test I wrote from some unit that I wanted to exist, deleting the implementation, walking them through a re-implementation, and then passing them the keyboard to implement another deleted implementation.

Make sure that the developers actually experience a cycle of red-green-refactor.

When you are selling high-priced products to consumers you have to sell an idea first and your product second. Let me offer an example.

I used to sell beds. Selling expensive beds to your every day person is really hard, because they are so expensive. Who is going to spend $7,500 for a luxury Tempur Pedic when they could get cheap mattresses from Big Co for $300? Eventually, I gave up trying to sell beds. It was just a waste of time because nobody every bought them.

Then something interesting happened: somebody gave me the advice I gave you.

You see, when you are selling something expensive, you can't just push the product. You have to sell your customer on the experience.

How did I apply this to mattress sales? It was easy!

Did you realize that you spend approximately 33 percent of your life sleeping? Sleeping is your body's way of regenerating itself. It's the biological equivalent of turning off your laptop when it starts to overheat. It gives your muscles and tissue time to rebuild themselves. It gives your brain time to digest everything you've learned that day.

A good set of mattresses will last you 20 years. With that said, very few people go so long without replacing their bed. In reality, a bed will last about 10 years; less if the customer has a lot of money or moves frequently. But let's assume a 10 year lifespan for your new $7,500 bed.

Over ten years, that bed will cost you $2.05/day.

Now is $7,500 a lot up-front? Absolutely. But like anything else of high quality, this bed is an investment. And what an investment it is. I will bet you that a luxury Tempur Pedic will give you a better ROI than most stock traders who have been in the biz for 20 years get.

How?

Simple. Better sleep = better productivity. If you got a great night sleep every single night, do you think you could turn that extra productivity into $10 dollars each day? I know I could! If I were productive for one extra hour each day, I would make an extra $100 a day.

Assuming you make an extra $10 a day from your new found productivity, this bed will pay for itself five times over throughout its lifespan.

And so on and so on.

You see, I'm no longer selling a bed. What I'm selling is a complete experience, an investment and a new way of life. $7,500 doesn't seem so bad when it makes your life better for the next 10 years now does it?

What's the point of my rambling?

You need to avoid selling your coworkers on testing. Very few people like testing and the only sure-fire way to make somebody do something is to make them WANT to do it.

Instead, try selling them on how their lives will improve if they increase the quality of their software. Show them how they will benefit from shorter release cycles, fewer bugs, higher quality code, etc. Show your coworkers the benefit of taking the time to correctly architect a solution before they start implementing it. How does taking the time to learn and correctly use the tools at their disposal benefit THEM?

Will this be difficult? Absolutely. People don't want to change. Your job is to make them WANT to be better. There is simply no other way to accomplish your goal with any lasting impact. Sell your team on the complete experience, the investment and the new way of life.

Hope my two cents helps. :)

"People don't buy what you do, they buy why you do it". You sell the bed not based on the quality of the materials so much as the improvement that it gives to their life.