Tech is only half the answer

Technology advancements continue to promise to streamline operations, cut costs, save time, reduce error rates, or unleash creativity.

Leaders hear these things and give their teams the charge to go find the technology that will produce these results. So they do.

And then, often, the tech collects dust, is either ignored or used less than expected, or leaders push for fast adoption and then wonder why they aren’t seeing the results promised.

You might think this is all about AI, but it isn’t. I’ve seen organizations bring in Salesforce and have it go unused by their sales teams because who enjoys logging conversations and data entry? I’ve seen project trackers and time trackers get introduced in an effort to…well, start tracking everything, only for them to go ignored, because there was no reason given for tracking everything other than “because we want to know.” I’ve seen 1:1 meeting tools and key performance indicator tracking systems be bought, introduced, adopted for two weeks, then dropped, because they caused more work than people thought they were worth, and nobody gave a good reason for continuing usage.

These tools cost money - often lots of money. And they all end up being only half of the answer to improving operations.

The other half of the answer is: change management, or working with people and processes to ensure the tech actually gets used and integrated into the organization.

First, the people. Nobody likes change. We are creatures of habit. We need a dang good reason to change out of our habits. And, “Because management said so,” is just as good as, “Because I said so,” to a child. Neither works.

One way to shift this is to position the change by showing what the benefit is to the individual. As someone once put it, we all tune into a single radio station: WII FM, What’s In It For Me? Julio in Accounting doesn’t actually care that the company is going to save $20M over the next four years by adopting the new processes and tech. He cares that he will get to go home earlier because he won’t have to manually add up the sums across multiple projects and enter it into a different system and triple check that all the numbers are correct…something that takes him a good 20 minutes per team, and he supports 5 teams!

Another way is to attach the changes to rewards and consequences. Let’s take that Salesforce usage as an example. If you tell your sales team that they need to keep Salesforce updated, but then you never bring up their accounts in Salesforce during their 1:1s to review account status, then they will never have a reason to keep Salesforce up to date! Or if your sales team starts working with other teams who are expecting data from Salesforce so they can do their jobs, they will now have a reason to keep Salesforce up to date - someone else is relying on them! If you now also tie a weekly shout-out to whomever has their accounts updated first at the end of the week, tapping into the competitive nature of salesfolks, you will see those accounts get updated!

Additionally, make sure your people have the right training to use the new tools. This doesn’t just mean having the user manual or tutorials that comes out of the box with the technology. This means having training that fits your team’s use cases and documenting the training, either using video or keeping written documentation. This helps the training scale and keeps it relevant.

And make sure that training takes place when your team has space to breathe, not when they are slammed with deadlines and other priorities. I learned this the hard way: I tried to schedule a workshop for a large department and was excited about the timing being in the spring, only to learn that the part of the department supporting the advertising group was in the middle of their busiest ad testing season! Delaying until everyone had time was inconvenient only to a few people; holding it when everyone could focus on the message mattered far more.

Something you may have noticed in the Salesforce examples: updating processes so that they involve the new tech tools. If a process remains untouched, it’s like saying the new tech is optional.

Don’t have your processes documented? That’s fine - now’s a great time to document them using either video or written documentation. It doesn’t have to be in a formal step by step format, but documenting your processes somewhere and showing where the new tech fits in is a necessary exercise.

And when it comes to AI, processes may need to be entirely rewritten and reconsidered! Don’t try to force fit your process to AI or the other way around. Rethink how the process should work with AI, then test it and document what works and tweak what doesn’t.

Tech, people, and process: together, they increase your chances for seeing a return on that tech investment.

A quick note before I dive in to the comparison between Quillit and Flowres: both of these platforms are constantly making improvements. What I’m sharing is based on my experience focused on uploading transcripts of in-depth interviews and analyzing them using the AI tools each platform provides.

For each, I uploaded 10 transcripts of in-depth interviews and used the same prompts and questions to analyze the transcripts.

The signup and getting started was super easy. Finding pricing required entering an email address, but it’s given once you provide that information.

For usability after signing in, uploading transcripts was simple. I could also upload a discussion guide and analysis plan to help guide the AI in the analysis, which was great! I could also upload demographics for the participant profiles. They also have pre-loaded prompt templates to help reduce hallucinations, which was very interesting to see. It took some of the question out of how to put the prompt together to ensure a good response.

Citations for the analysis was always on, and the analysis was strong overall. The summary analysis here comes first, and then you can ask follow-up questions without using the pre-loaded prompts if you want. I did find the output to be a bit more terse than Quillit’s output.

The drawback for me was in the UI, where I found a couple of items confusing (like where to find the report after not using the tool for a couple of weeks), and the lack of export for the analysis. It also applied a theme to a single quote, which, while the quote was relevant, identifying theme based on just one quote might not be entirely appropriate.

In fairness to them, I’d let them know about the issues I’d faced, and they already updated the UI so that the transcript analysis button is super easy to find, and they shared that the button by the title of the conversation of the Q&A for the analysis is where the export feature was. So, kudos to them for quickly acting on feedback!

Quillit was also extremely easy to sign up for and get started. It was easy to start a project and upload transcripts. Similar to Flowres, I could upload a discussion guide to help guide the AI in the analysis. I could also upload demographics about my participants to help with the analysis.

With Quillit, there were no pre-loaded prompts. Instead, you enter the business context and research objectives and then have space to enter specific questions that you want addressed in the analysis. This is different from Flowres, where the transcripts were coded and then you went into a Q&A style screen to start analyzing the data. There is information to help you with what to include in the business context so that you get the best output possible.

The analysis itself was more verbose, and there was a toggle for including citations or not in the report. Similar to Flowres, I couldn’t export the analysis but had to copy and paste if I wanted to capture it in a document.

The drawbacks here were that there was an instance of a theme being given to a single quote, though I tried asking it for multiple citations per finding. I also didn’t think they had an export function, similar to Flowres, until I reached out. Turns out they do, and they let me know where to find the function (three dots on the tile of the saved report has the download option), and they’ll be looking into making that easier to find as they work on updates to their tools.

Honestly, I think both platforms have done a great job mitigating where possible for hallucinations and are pretty dang robust for getting you started on analysis of qual studies. The price points seem pretty nice, and they are both very responsive and supportive of small businesses and solopreneurs.

Jiten and Annie are also both wonderful people who are super responsive to their customers, supportive of small businesses and solopreneurs, and take feedback about their systems seriously. I would recommend looking into either option for your qualitative research analysis!

Trigger warning: discussion of depression symptoms.

Depression is a very difficult thing for many who have never experienced it to understand. It’s invisible, and it’s silent. It affects 3.5% of the global population. And I think that’s an under-representation, since that number relies on self-reporting/

But here’s how it impacts me, personally. I will have days that are going along just fine, and then, one day, I wake up, and I just…feel off. I don’t feel sad, necessarily, I just feel like I can’t smile or laugh as readily as usual. I feel sluggish - my reaction times feel slow, my thoughts feel slow, and my ability to comprehend what’s happening around me feels slow. And then it lingers. No amount of exercise or healthy eating or social interaction seems to have a lasting effect. So, instead of thriving, I move to surviving. And then my anxiety kicks in and sets the worry meter on, “This will never get better.”

It’s like waking up after many brightly lit days to a day that is cloudy. You notice the clouds, but you don’t notice the bank of larger, darker clouds rolling in. The next day, the clouds are a little heavier. The next day, a little heavier still. The lights dims a little more until you find yourself in the dark, and the 10-day forecast continually says that’s just how the weather is going to be, and you start to forget what the brightly lit days felt like, and you start to think you’ll never feel the sun again.

Except then, you also start to think you’re the one who somehow made it cloudy and overcast, because everywhere you go, it’s the same. You can’t find a vacation escape that isn’t cloudy and overcast and dark. And so you start to stay in bed longer, because what’s the point of self-care if there’s no reason to leave the house? And you start to retract because you’re sure nobody else will understand how you feel, because when you do leave the house, everyone else seems to be acting like it’s still sunny, so it must just be you that’s seeing and feeling the dark.

You see how someone in that state ends up just feeling broken, and how depression lies, and how not talking about mental health makes people feel like they’re alone when their mental health is low?

Right now, in the US, the economic situation is difficult and is impacting small businesses heavily. I can only suspect many might find themselves experiencing depression, unsure where their laughter went. So, if you haven’t heard from a friend or if you notice a colleague smiling a little less, check in on them and offer a friendly ear to listen.

Mental Health Crisis Hotline: https://988lifeline.org/