Why I Chose Not to Automate 21 Recruitment Touchpoint

One step that I feel should not be automated is the call a recruiter has with the candidate after an interview with a hiring manager. With new technologies, there are a lot of parts in the recruiting process that can be automated, such as scheduling and screening, but this part in the process warrants human intervention. It's important that recruiters take the time to align with the candidate and ensure they still feel good about the position and want to move forward. Job seekers spend a good deal of energy and time in the recruitment process, so taking a moment to genuinely align with them can help them feel more valued and engaged. Doing this can foster a stronger candidate experience, as well as bolster a company's employer brand, because that human connection might make the job seeker still want to pursue opportunities with the company if not selected for the role.

The bit I keep firmly human is judging someone's actual work, the paid trial task, and the real conversation around it. No automated screen can tell me the thing I most need to know.

Plenty of the rest gets automated now: resume filtering, keyword matching, the initial sort, and fair enough. But the moment of genuinely deciding whether a person can do the work and fit the team stays with a human, always.

It works in our favour because the automated tools are precisely what everyone games. A perfect-on-paper candidate tells me almost nothing, while watching someone actually tackle real work tells me everything. It also respects the person, since genuine ability deserves a human's judgment, not an algorithm's verdict. Keeping that evaluation human is a big part of why our hires actually stick. A machine can filter a resume all day, but it can't spot the person who's going to genuinely thrive on the team.

We also chose not to automate feedback for creative and technical talent on the app platform and in hiring agency personnel. When an applicant completes a design mock-up or technical assignment, a senior lead does a real-time review of that work with them. The senior lead will go through each element of the completed assignments with the applicant and share thoughts about how the applicant's decisions were made, what they liked/liked least about it and other things.

Automated systems do not capture creative intent, problem solving ability or an individual artist's unique approach. It is our experience that interactive discussion around portfolio work shows our agency's focus on collaboration and respect for the creative process. The candidate also expresses gratitude for the thoughtful feedback which helps us continue to attract top-tier creative talent, and creates a positive perception of our company as a supportive environment for such talent.

We made a conscious decision to avoid automating the process of total compensation evaluations for executive and administrative positions. Prior to producing formal digitally generated offer letters, an employee services manager will personally meet with each candidate to discuss how they are paid, what is included in their health insurance coverage, and how we match their contributions to a retirement plan.

This transparency of financial information helps eliminate potential misconceptions about the overall "total reward" of working at this institution. The ability for candidates to have their questions answered by the individual responsible for explaining all aspects of their employment benefits further builds on the concept of institutional trust and provides evidence that we value being completely open and honest about how employees receive compensation.

We have chosen to exclude an applicant tracking system, which would be used as an applicant filter using keyword parsing, from our screening process for administrative support positions. Keyword-based ATS systems often automatically disqualify excellent job prospects with no specific buzz words in their resume, even though they may possess great experience in office related work.

Our talent acquisition team member will personally examine each of the resumes received. The manual examination of resumes by humans provides the ability to find alternative background experiences, veteran status, and skills transferable to other jobs which can be easily missed by the algorithms developed within ATS systems. Using a human method of examining resumes for qualifications increases the number of candidates who qualify for interviews; it also promotes the diversity of our workforce and enables us to assess candidates on their total professional abilities and not solely on how well their resume matches certain pre-determined search terms.

We decided on purpose to refrain from automating the process of delivering job offers for administrative and operational roles. Instead of having the hiring manager send an automated email notification that includes an e-signature link for the candidate to electronically sign and confirm, we have the hiring manager call the candidate and extend the offer verbally, as well as explain the responsibilities associated with the position and how excited they are to welcome them to their new team.

The personal connection at this point in time establishes a very positive, welcoming environment prior to the company generating formal paperwork. This provides the candidate an opportunity immediately after receiving the verbal offer to inquire as to when they can expect to begin employment, what benefits will be provided, as well as who will be working within their new department. Delivering job offers personally has increased our overall rate of accepting job offers from candidates because it reinforces candidates' decisions to accept an offer and become part of our organization by making candidates truly feel welcomed and valued based upon their qualifications.

The rejection. Everything upstream of it runs on rails: scheduling, acknowledgements, the first structured screen. But when we say no to someone, I write it, and I say the actual reason in one sentence. It costs me maybe four minutes per candidate. The reason is not sentiment; it is signal. A templated rejection teaches you nothing, while a specific one gets replies, and those replies are the only honest feedback we get about how our own brief reads from the outside. Two people we turned down came back the following year, and one of them we hired. Neither would have if the last thing they heard from us had been a mail merge. Automate the parts of hiring where being fast is the kindness. Keep the parts where being specific is.

One touchpoint we deliberately keep human is client calibration at the outset of a specialist search across crypto finance, compliance, growth or leadership.

For us, that is where the quality of the appointment starts. We want to understand the business context for the hire, the change the organisation is trying to effect, how the role sits within the operating model and what success will look like in practice.

That means looking beyond the job description to accountabilities, decision-making authority, organisational interfaces, reward, succession and the capability already present within the team. It gives us a much clearer view of the role the business actually needs to hire for.

Good calibration also helps establish what is genuinely essential and what is simply conventional. That creates a stronger brief, a broader and more relevant market, and better assessment criteria.

It has a direct impact on candidate engagement too. We can represent the opportunity with real clarity, answer senior candidates' questions properly and give them a credible view of the organisation, the role and the expectations around it before they enter a process.

Technology is extremely useful for research, market mapping and process efficiency. We use it extensively. But the work of defining the appointment, aligning the stakeholders and setting the parameters for a successful search is something we deliberately keep human.

Our administrative staff positions were selected intentionally to exclude automation of the first round of candidate screening calls. Many companies use automated video interviews or online surveys as an initial screening method; however, we have chosen to conduct a live 15-minute telephone interview with each candidate via a recruitment representative. The live phone interview provides us with the opportunity to build rapport and create a relationship of trust with each candidate, beginning immediately upon their first contact. In addition, a live conversation also permits us to evaluate candidates' communication skills, provide answers to candidates' questions in real-time, and convey our company's values. Candidates have reported that having a "live" person involved in the hiring process from the onset has created a sense of being "valued," versus feeling like they are simply "processed." Maintaining a human element at this point in the hiring process results in increased levels of candidate engagement and ultimately reduces dropout rates during the application process. By providing the ability to select the most qualified administrative staff candidates, while competing in today's job market, we can ensure that our staffing needs will be met effectively.

It's the interview. I know a lot of AI recruiting tools are pushing AI interviewers because they say it's more objective, there's no gut feeling, and all of that. But I think interviewing is one thing you just can't automate.

An interview is a conversation. It's not just a checklist of, "Do you have experience here? Do you have experience there?" You have to be able to listen to the candidate's answer and ask follow-up questions based on what they said. You have to dig deeper and pressure-test their experience to really understand if they actually know what they're talking about.

And it's also not just about experience. You're trying to understand how they communicate, how they think, how they explain things, and how they work with people. I don't think AI can accurately evaluate all of that yet.

I also think pushing AI interviews too far can dehumanize the recruiting process. If you're a candidate and you go through an entire hiring process without actually talking to a person, what does that tell you about the organization you're potentially joining?

AI has a place in recruiting, but it should support the process. It can help build a candidate scorecard, create a job description, transcribe interviews, and take meeting notes. Those things save recruiters a lot of time. But when it comes to actually interviewing and vetting a candidate, I still think that needs to be human.

400 or so applications came in for a single role last quarter and every rejection was still written by a person. Not a template with a name merged into it. Someone reads the CV again, writes 2 lines about what was missing and sends it. It takes our recruiter most of a Friday.

Everyone here logs in from a different city, all 60 of us, so the only thing a candidate ever meets is text on a screen. A rejection is the last thing most of them read from us and it is the one part of the process where there is nothing to gain. Automating it would save roughly a day a month. You have been on the other side of the instant rejection that lands 90 seconds after you applied.

The final interview before someone starts cleaning turnovers alone stays human. Everything else is automated: the application, the background check. That part should move fast. But before someone gets a key code and an empty house, I still sit down and talk with them myself. I ask how they'd handle finding something broken, or a guest's things left out on a counter. The answer tells me more than any checklist. Only 1 in 60 applicants make it through our hiring process, and I decide most of that right there in the room. A background check confirms a record is clean. It doesn't tell me if I'd trust someone alone in a stranger's home with nobody watching. That's a judgment call. Keeping it human slows hiring down some weeks. I still won't automate it. Turnover cleaning runs on access to somebody else's house while they're not in it.

The one touchpoint I have chosen not to automate is the final decision meeting with internal interviewers before extending an offer. Software can collect ratings, but numbers alone flatten the story. A live discussion forces people to defend impressions, compare notes, and separate evidence from enthusiasm or bias.

That choice improves talent acquisition because the best hiring decisions usually come from careful calibration, not averaged scorecards. One interviewer may have caught coachability, another may have noticed inconsistency, and both details matter. A human debrief creates room for those observations to connect. It sharpens standards, reduces snap judgments, and leads to more coherent yes decisions. Candidates hired through that process tend to match the role more closely and ramp with fewer surprises.

The first real conversation stays human, and it is done by whoever the person will report to. Not by me, and not by a form.

We automate the parts around it happily. Scheduling, acknowledgements, keeping candidates informed about where they are, all of that is better done reliably by software than sporadically by a busy person. What does not get automated is the 40 minute call where somebody hears what the job is like on a bad week.

That call is not a screening exercise. It is where I describe the fortnight before Christmas honestly. The volume, the pressure, the fact that everyone packs boxes when it gets deep. Roughly one in five people quietly step back at that point, and that is the call doing its job. The ones who stay have chosen the real version of the role.

The benefit shows up in retention rather than in speed of hiring. Somebody who took the job knowing the worst of it does not resign in January feeling misled.

A candidate once asked me at the end of that call what I found hardest about the business. Answering honestly turned it from an interview into two people working out whether this was a good idea. She has been with us for years.

While we could have relied upon the automated communications available in our onboarding portal to automatically send reminders to all new hires one week prior to their first day at work, instead of this process we made the decision to allow the direct supervisor to make contact with the new hire via telephone in order to extend a welcome and also provide answers to any logistical or operational issues related to the employee's first day.

A very common reason why many new hires experience "first day jitters" is that they are unclear about what to expect on their first day. Allowing supervisors to personally reach out to each new hire and establish open lines of communication prior to them even setting foot in the workplace greatly assists in reducing those feelings of uncertainty and anxiety associated with first days. In addition, it provides new employees with a sense of being supported and informed, as well as developing a connection with their manager from the onset. A welcome call made by management will help reduce early employee turnover, create trust between management and employees, and set the stage for long-term engagement.

From a management perspective, we made a conscious decision not to incorporate automation into team Q&A sessions that occur as part of the final interview process for facility support staff. We reserve specific times in which candidates are able to have private conversations with other potential employees who will be working at the same facility but will not have management involved.

This allows us to create a real sense of authenticity by providing our current and future staff members the opportunity to freely communicate the details of how work is done on a daily basis, how workload is distributed, and how teams interact. Our current staff members also get to help shape their department through these interactions. In addition to building camaraderie among new and existing support staff, this process provides candidates with a true understanding of the daily operational processes of the facility they would be employed at. This ultimately helps with the transition into their job when they begin employment.

“The one step we deliberately kept human is deciding which candidates the client actually sees.

Our AI matching tool scans our whole candidate database in seconds, which used to take a hunter hours, so I’m not against automating that part. What it hands back is a set of strong matches and then a middle group it flags for human review, because it can’t tell yet. We never let it discard that middle group, and a good share of the people we place come out of it.

The tool is comparing a resume against a brief. It doesn’t know what stage the company is at, or that this hire will be the first person in that function, so a recruiter who was on the kick-off call can look at someone missing two or three of the nice-to-haves and say this is the right person for that stage. Sometimes we give the client both options: someone who ticks every box and sits above their budget, and someone missing a couple of things we don’t think are critical who fits it. They choose.

On more than 70% of our searches, there’s a recruiter and a hunter working it together, and the recruiter is the last filter before anything reaches the client. A high match score on its own isn’t a reason to send someone.”

Maria Jose Lorenzo, Director of Talent Acquisition, Hire With Near

Maria Jose Lorenzo is Director of Talent Acquisition at Hire With Near, where she leads a recruiting team of more than 40 people placing Latin American professionals into remote roles with U.S. companies.

The welcome moment for a new hire. We automate the trigger, never the message.

Our platform tracks the start date and gets a gift to the person's desk or door before day one. What we will not automate is the note that comes with it. A hiring manager writes it, by hand, and names the specific thing that made them want this person on the team. It takes four minutes.

The reason is simple. A new hire can tell the difference between a company that has a process and a company that was paying attention. The automated part protects the logistics so nothing gets missed. The human part is what the person actually remembers, and it is the first proof that the culture described in the interview is real. Candidates talk to each other. That first-week story travels.

One touchpoint I would never automate is the in-depth conversation before we shortlist a candidate. A CV can tell you whether someone has the right experience, but it cannot show how a family communicates, what makes them feel comfortable, or whether a candidate's personality will genuinely fit the household. Those conversations often reveal the small details that determine whether a placement lasts, so I see automation as useful for administration, not for replacing human judgment and connection.

The one step we never automated is the alignment conversation before anyone joins. That's the final conversation after technical interviews, culture checks, and reference calls. It's where we talk through what we're building, why we're building it, what the constraints are, and what conviction actually means when the team is three people.

Most crypto companies run recruitment through playbooks optimized for scale. They hire fast, onboard faster, and burn through people when the token cycle shifts. That model works when you're staffing for a phase rather than building for permanence. When your team is three people and you're shipping five product lines without a marketing function, every person either multiplies output or fractures it. There's no middle layer to absorb a bad fit.

The alignment conversation surfaces what standard interviews miss. Technical ability doesn't predict conviction. Culture fit doesn't predict what someone does when the work gets harder than the interview suggested. I care about whether this person understands that we're not raising another round to hire around problems. I care about whether they've thought through what it means to work on something where the feedback loop from user report to shipped fix is measured in days, not quarters, because the decision-making layer and the execution layer are the same layer.

We've turned down people who interviewed well technically but framed the opportunity as a stepping stone. We've turned down people who wanted more clarity on the token launch timeline than on the product roadmap. That clarity only comes from a conversation, not a rubric.

Automating that step would optimize for speed. We optimized for conviction instead. Every person on this team is here because they chose to be here after we told them exactly what this looks like. That's the only hiring moat a three-person team has.

The one human touchpoint preserved is reference conversations that go beyond title verification. In many hiring systems, references are reduced to a compliance box, often automated, standardized, and shallow. For a small company where each hire meaningfully affects execution quality, that approach leaves too much unknown. The goal is not character theater. It is understanding how a person actually behaves when priorities collide, feedback is uncomfortable, or operating conditions are messy.

That choice improves talent acquisition because references often expose pattern consistency, not surface charm. I ask about follow-through, precision, receptiveness to evidence, and whether trust increased or decreased over time. We have learned that long-term fit is usually clearer in those specifics than in polished interviews.