My first job in tech was to make sure that lists of words in any language alphabetized properly on the computer when the user asked them to. It sounds as simple as ABC, but when you’re handling strings like ඇපල්, ആപ്പിൾ, and ᑭᒻᒥᓇᐅᔭᖅ — often in languages without any agreed-upon standards — things get complicated fast.
My boss, who had done the job before me, was a linguist with expertise in complex writing systems. On average, it took her two months of manual work with paper dictionaries to define the sorting system for a new language, and another few weeks for an engineer to implement her analysis.
I didn’t have any particular background in writing systems. On the other hand, I knew how to code, and I saw a way to make the work go faster. Rather than working with paper dictionaries to figure out how a new language worked, I was able to code up user tasks for native speakers to tell me which words came in which order, and to automatically extract the right information from their opinions. The new approach cut down our average language implementation time from three months to one week. 12 times faster!
I saw first-hand how my ability to automate a task with code made me more effective. I turned a conventionally offline job into an engineering job, just by virtue of seeing what could be automated.
Fast forward twelve years.
Job automation talk is everywhere. Will AI replace us? Will robots make it hard for people to find work? If bots take jobs, should they pay taxes? Do we need a Universal Basic Income to offset the impact of automation in our employment market?
While these are all significant questions, they miss two critical points. First, many traditional enterprise jobs have already been reinvented by automation. Second, the automation revolution is not always forced on workers from above. Job automation often starts with someone on the job who recognizes an opportunity for automation to make their tedious tasks go faster — a lot like me in my first tech job.
Customer success roles provide a great example. To get hired doing customer success at Textio and most other SaaS companies, you need to shine in two main areas: You need to be able to build relationships with people that help them succeed over time, and you need to be able to weave data from a broad range of sources into a coherent narrative.
But increasingly, we also look for people with another skill that supports this foundation: You are more effective if you can find interesting patterns in large data sets without the help of an additional engineer. In our line of work, finding interesting patterns means that you have at least basic scripting ability. With the growing ubiquity of machine learning in all kinds of software products, I’d wager than within five years this becomes a required skill set for the role — not just at Textio, but everywhere.
The reinvention of customer success is not unique. Work across the enterprise has become increasingly technical. It is no longer unusual to find project managers creating complex Excel macros, ops professionals building tools to automate the collection of sales data, or user researchers writing SQL queries to understand large data sets. For these and many other roles, people who can combine a solid technical foundation with traditional skill sets are more successful than people with the traditional skill sets alone.
Even where coding as such is not required, many enterprise workers are now expected to use predictive automation technology as a part of doing their best work. Sales leaders who collect and respond to deep and specific data about what has worked in the past are better sellers. Marketers with the tools to quickly test and evaluate collateral performance can make adjustments faster. And as we’ve seen at Textio, hiring teams that use our enormous data set to craft their job posts fill roles almost 20% faster, with nearly 15% more people applying from underrepresented groups.
What is happening in many of these cases is not job replacement, but job reinvention: people are augmenting their traditional skill sets with automation to get deeper data insights, faster product throughput, and operational improvement. Companies that work with automation, data, and learning loops simply outperform companies that do not.
Despite the prevalence of automation in job reinvention, the ability to work with people has never been more critical. Go back to that customer success engineering role for a minute. Data is only as good as the context surrounding it. Telling stories with data, and even figuring out which data to collect in the first place, is a fundamentally human task. Our customer success engineer is as served by her background in psychology, linguistics, or business as she is by her scripting ability.
Look to the bottlenecks in your work and I’ll show you automation coming faster than you think. It just takes someone already on the job to recognize the opportunity.
Learn more about how language impacts your hiring at textio.com