Purposeful and realistic return on investment calculations

It's a time of the year that allows us all an excuse to message people we usually don't. I had a group like that in mind: managers of all of my tester colleagues at work. There's two reasons why I usually don't message them:

1. We coexist in the same pool as peers, but we have no official connection 
2. There is no list of them easily accessible

You can't create a connection with your peers without being intentional about it. Having now tracked through coverage of fellow customer contact people (200+ by now), I needed a new dimension to track. This is just one of them, although an important one. 

Yet still, you can't contact them if you don't know who they are. With large number of test professionals, asking around won't work. And the manual approach of tracking through organization chart is probably a lot of work, even when the data is available. 

For six months, I have accepted that such list does not exist. Call it holiday spirit or something, but that response was no longer acceptable to me. So I set out to explore what the answer is. 

Writing a script can be one of those infamous "10 minute tasks", but for purposes of discussing return on investment, let's say it takes 300 minutes. Maybe it is because originally I wanted to create a list of test professionals and automatically update it to our email list because I hate excluding new joiners, but since I am exploring, I never forgot what I set out to do, but I may have ended up with more than I had in mind at first. I take pride in my brand of discipline, which is optimizing for results and learning, not to a commitment made at a time when I knew the least -- anything past this moment in time. 

So, I invested 300 minutes to replace 299 minutes of manual work. Now with that investment, I am able to create my two lists (test professionals and their managers) investing 1 minute of attention on running the script (and entering 2FA response). How much time is this saving, really?

I could say I would run this script now monthly, because that is the cadence of people joining. Honestly though, quarterly updates are much more likely because of realities of the world. I will work with the optimistic schedule though, because 12x299 is so much more than 4x299. Purposeful return on investment calculations for the win! 

You may have done this too. You take the formula: 

You enter in your values and do the math (obvs. with a tool that creates you a perfect illustration to screenshot). 

And there you have it. 1096% return on investment, it was so worth it. 

The managers got their electronic Christmas card. I got their attention to congratulate them on being caretakers of brilliant group of professionals. I could sneak in whatever message I had in mind. Comparing cost and savings like this, purposeful investment calculations, its such a cheap trick. 

It does not really address the value. I may feel the value of 300 minutes today, but is there a real repeat value? It remains to be seen. 

That is not all that we purposefully do with our investment calculation stories. I also failed to introduce you to the three things I did not do. 

1. Automating authentication. 2FA is mean. So I left it manual. Within this week we have been looking at 2FA on two Macs where automating it works on one and not on other. I was not ready to explore that rabbit hole right now. Without it, the script is manual even if it runs with 1 minute attended and 15 minutes unattended. 
2. Automating triggering it with a pipeline. It's on my machine. It needs me to start it. While pushing the code to a repo would give access to others, telling them about it is another thing. And while it works on my machine, explaining how to make it work on their machine given the differences of machines and knowledge levels of the relevant people, this should really be just running on a cadence. 
3. Maintainability. While this all makes sense to me today, I am not certain the future me will appreciate how I left things. No readme. Decent structure. Self-documenting choices of selectors, mostly. But a mental list of things I would do better if I was to leave this behind. I would be dishonest saying this is what I mean as work when I set out to leave automation behind. 

When you factor all of that in, there is probably a day or a week of work more. Week because I have no access to pipelines for these kinds of things, and no one can estimate how quickly I will navigate through the necessary relationship building to find a budget (because cloud costs money), gain access and the sort. 

The extra day brings my ROI to 380%. A week extra brings me to: 

Now that's a realistic return on investment calculation, even if it still misses the idea that this would could be not done, not repeated if it was not valuable. And the jury is still out on that.

This story is, most definitely, inspired by facing the balance of purposeful and realistic often for work. And I commit to realistic. Matter of fact, I am creating myself a better approach to realistic, because automation is great but false expectations just set us out to failure. 

A Six Month Review

That time of the year when you do reflections is at hand. It's not the end of year yet, even if that is close. But a few relevant events in cadence of reflections culminated yesterday and today, leading me to writing for a seasoned tester's crystal ball. 

The three relevant events are all worth a small celebration: 

• Yesterday was first day after trial period at CGI as Director specializing in testing services and AI in application testing. We are both still very happy with each other. I should say it more clearly - I love the work CGI built for me, and I love that I get to bring in more people to do that work with me also next year. CGI's value-based positioning as an organization that supports volunteering and great community of testing professionals (testers, developers, product owners and managers/directors throughout the organization showing up for testing) has been a treat. 
• Today the TiVi ICT 100 Most Influential list was published just for Finnish Independence Day, and I found my name on the list for 6th year in a row. You can imagine there are more than 100 brilliant professionals influencing in the ICT in Finland, and Finland has had a good reputation of educated ICT professionals with internationally competitive prices meaning that for a small country, we have a lot of brilliant people in ICT. Representing testing on this list even with title 'Director', is a recognition for all the great people teaching each other continuously in the community. Testing belongs. 
• Today we received positive news of one frame agreement bid I had been contributing to in my 6 months at CGI. Building a series of successes in continuing as a partner of choice with me around brings me joy. 

While these things are the triggers of reflection today, it is also a good time to take a look at themes of testing coming my way. 

Distributing Future Evenly

When searching for work with a purpose, I welcomed the opportunity to put together themes of relevance: 

• Impact at Scale. I believe we need to find ways of moving many of our projects to better places for quality and productivity. Many organizations share similar problems and there has to be ways of creating common improvement roadmaps, stepping plans to get through the changes, and seeing software development (including testing) with increased successes. Scale means across organizations, but also over time when people change. 
• Software of Relevance. I wanted to work on software that I feel connection with. I have been awarded multiple customers with purposes of relevance and feel grateful for the opportunities. 

Turns out this is a combination of conversations internally with people I work with, our current and potential customers, and the testers community at large. 

I have met over 150 peers at CGI in 1-on-1 settings (I tracked statistics for first four months and reacting 160, I decided counting something else would make sense). I have enjoyed showing up with our internal Testing Community of Practice in Finland, and started to create those internal connections globally. I have learned to love the networked communication expectation where hierarchy plays a small role. I've done monthly sessions for external testing community of practice Ohjelmistotestaus ry with broadcast of topics I work on. And I have been invited to conversations with many clients. 

AI in Application Testing

With testing as my backdrop living up to a reputation of 'walking testing dictionary', there is a big change into the future we have with increased abilities in automation with AI helping with quality and productivity.

I have looked at tens of tools with AI in them. For unit testing, for test automation programming, and for any and all tasks in the space of testing. From exploring these tools, I conclude a need of seeing through the vocabulary to make smart choices. With models as a service, it's what we build around those models that matter. The technical guardrails for filtering inputs and outputs. The logic of optimizing flows. The integrations that intertwine AI to where we are already working. 

In these 6 months, I created a course I teach with Tivia ry - 'Apply AI of Today on Your Testing'. Tivia sets up public classroom versions, and both they and CGI directly are happy to set up organization specific ones. 

The things that I personally find exciting in this space in the last six months are: 

• Hosted models. Setting up possibility to have models on your own servers (ahem, personal computers too) so that I can let go of modeling what my data tell about me or reveals from my work. 
• Open source progress. Be it Hercules for agents turning Gherkin to test results, or any of the many libraries in the Selenium ecosystem showing proofs of concepts on what integrations are like, those are invaluable. 
• Customer Zero access. Having CGI be a product company allows us to apply things. Combine the motivation and means, and I have learned a lot. And yes, CGI has a lot of products. Including CGI NAVI, which is software development artifact generation product. 

Digital Legacy

The time focusing on scaling testing has warranted me revisiting my digital legacy. From being able to create AI-Maaret by having 20 years of my thinking on written formats, to making choices of how I would  increase the access to my materials for everyone's benefit has been a continuous theme. 

From 50 labels of courses in Software Testing, I brought things down to only 18. 

Should testing expertise be a capability you are building, we could help with that. We're currently prioritizing which of these we teach for CGI testing professionals next year. In addition to all the other training material we already have had without my digital legacy. 

Purpose-wise, I am now moving from having access to this myself to institutionalizing the access. And that means changing license for next generations of this from CC-BY to CC-BY-NC-SA, and you could always purchase licenses that allow other uses too. 

Collaboration and Sharing

With tentacles in the network towards all kinds of parties, I still facilitate collaboration. My personal vision is frame a lot of work with open sharing, this being my current favored quote.

The tide is something we create together. And we want to go far. Sharing openly is a platform for that collaboration so you see me sharing on Linkedin, Mastodon and now also Bluesky. My DMs on these are open. 

Networking

With open DMs and email (maaret.pyhajarvi (at) cgi . com), I am available for conversations. I have targets set on networking and should you have a thing we can discuss for seeking mutual benefits, don't be a stranger and get in touch. 

I am open to showing up in events, but I am still not proposing talks in CfPs. Please pull if I can help, and I help where I can. 

I show up for Tivia ry to take forward the agenda of Software in Finland. I show up for Ohjelmistotestaus ry to build a cross-company platform on testing. I show up for Mimmit Koodaa in any way I can, because that program is the best thing we've had for many many years. And I show up for Selenium, as volunteer in the Selenium Project Leadership Committee. I show up for FroGSConf open space as participant because it is brilliant. I show up for private benchmarking, currently with three groups of seasoned, brilliant professionals driving a global change in testing through supporting one another. 

As new thing, I have discovered "Johtoryhmä", women in ICT leadership in Finland. Will show up for that too. 

And when we meet, please talk to me. I am still socially awkward extrovert who imagines people don't  automatically want to talk to me. You talking to me helps me. And our conversations may help us both.