Atlassian Recruitment: November 2010 update
Australian international software company, Atlassian , have provided me with plenty to comment on in the past 15 months.
On 31 August 2009, I first wrote (see my blog) about the very different tactics Atlassian were adopting for a specific campaign they had just launched to hire 32 engineers for their Sydney head office.
This post generated 39 comments and, subsequently, prompted a story in the Australian Financial Review about the Atlassian campaign and the response from the recruitment agency world.
I wrote follow-up pieces on 4 November 2009 and 23 February 2010.
The impact of Atlassian’s hiring and retention strategies have been noticed (and recognised) far beyond the Australian recruitment agency world. Here’s a roll-call of their recent successes
- 1 of the 25 Best Small & Medium Companies (USA)
- HR Leader Highly Commended Award for Employer of the Decade (Australia)
- HR Leader Highly Commended Award for Employer of Choice (Australia)
- HR Leader Award for Innovation in Recruitment and Retention (Australia)
- Fairfax Employment Marketing Award for Best On-Site HR Team (Australia)
- a 2009 Hewitt Best Employer Highly Commended Award (Australia)
Now Atlassian are embarking upon the next phase of their expansion and have released the terms of engagement for any recruitment agency wanting to refer candidates to them.
To save you scouring the whole document here is my Reader Digest summary of the relevant bits:
-
Atlassian have abandoned their fixed fee for offers within specified salary ranges (which translated into fees in a range of 10.8% to 12.8% of salary) in favour of a more traditional fee of a percentage of starting salary-plus-super.
-
Recruiters can submit a maximum of 4 candidates but the more candidates you choose to submit the lower the fee you must agree to if any of your candidates are hired (ie 1 candidate submitted = 18%* fee, 2 = 17%, 3 = 16%, 4 = 15%) * denotes including GST
-
Unlike the original Atlassian32 terms, recruiters must choose how many candidates they will submit at the time of signing the terms of engagement. If you choose to submit one candidate, then a week later you source an even better candidate – tough luck, you can’t submit them.
-
As was the case previously, a full three month money-back guarantee applies to all placements
-
As long as you are happy to sign the Atlassian terms then ANY agency can submit candidates
There’s plenty of vacancies available at Atlassian. Currently there are vacancies in 15 different job categories within the Atlassian Sydney head office but unless you specialise in placing software engineers (or closely related) then you’re unlikely to get too excited by looking at the vacancy list.
Plenty of recruiters have expressed their unhappiness with these terms and have chosen not to submit any candidates.
Playing the devil’s advocate, I see the traditional alternative to Atlassian’s terms of engagement with recruiters as far less appealing. This alternative is the dreaded tender/PSA submission.
Preparing PSA submissions are inferior to the Atlassian terms because:
- They take a huge amount of time to research and complete
- You are required to provide a significant amount of confidential information.
- The process from start to ‘winners announced’ typically takes months.
- If you lack a track record with the client then the chances of being selected are slim
- Large agencies have an advantage because they have the resources to respond to tenders more effectively and they appeal to risk-averse clients (the nobody-got-fired-for-choosing-IBM principle)
- Your skill in writing winning tender submissions may not be a reflection of your skill in successfully filling clients’ jobs
- In almost all cases, the lowest price wins
None of these factors come into play at Atlassian.
It’s very straight forward – if you have the candidate(s) Atlassian want, then you win.
I’m a fan of keeping recruitment decisions out of the hands of tender writers, procurement departments and bureaucracies. What about you?
hi Ross, great post.
We basically let recruitment companies select their own rates based on how much work they take off our shoulders.
We've made the change after last year's campaign, and and I think it's the easiest and best deal in the market. Basically, recruiters that deliver the quality candidates will win out of this. I wouldn't know how to make it any fairer for both recruiter and the company.
All I'd say is that, intuitively (as a recruiter), the requirement for specifying in advance the number of candidates you're going to submit doesn't necessarily guarantee that Atlassian will get the best candidates on the market.
Here's my thinking: as a rational recruiter (ie. one acting in the most sensible way), I'm going to always get the best candidates I come across to register with me exclusively, and I'm going to make sure I place them with the clients who are going to pay me the most.
I'm also going to specify that I'll submit only one candidate to Atlassian (presumably my best current exclusive candidate at the time), in order to get the highest fee if successful.
So I submit that one, and a week later, register an even better one.
See where I'm going with this? I couldn't submit him/her to Atlassian even if I wanted to (which I don't anyway, because I'd be accepting a lower fee) so I'll just place him/her with another client (maybe one of Atlassian's competitors?) and, hopefully, charge them higher than 18%.
I'm happy, the candidate's happy, Atlassian have lost out. Any thoughts?
(BTW I'm not an IT recruiter and never have been, so maybe IT is a different market to the one I'm used to. But talking to colleagues who do recruit in IT, I don't think it is…)
Atlassian's model is surely innovative and has several advantages over some of the traditional tender processes.
It is fair and it opens up the door to good talent but I think it can be enhanced to become even better.
The key here is Atlassian is not the only company out there that wants to hire "good programmers". Their competitors will also accept nothing less. Their competitor may pay a headhunter higher fees to find that good candidate and promote their company to ensure the candidate joins them. The headhunter on the other hand can potentially treat Atlassian as the second best. It works for the headhunter, not Atlassian and problem is it all happens behind the scenes.
This model in my opinion can be enhanced by starting to treat each case in a more unique fashion. For instance by categorizing roles in order of urgency and rarity of experience. The rarer and more urgent the role, the higher the fee to find the right candidate.
Keeping the door open is great but also it may work best to build relationships with search oriented organizations which may have slightly higher fees but are faster and more reliable.
Great post Ross and I completely agree that any kind of innovation in the recruitment process that avoids the cumbersome PSA process is to be eagerly embraced.
I can see what Atlassian are seeking to achieve and I applaud them and wish them luck. I'm not sure that making recruiters choose how many candidates they will submit beforehand is quite striking the right note though.
I also recruit in a candidate-driven sector, similar to IT, and my access to quality talent constantly ebbs and flows depending on a whole range of factors coming out of referrals, advertising, networking, seasonal factors, and, yes, luck too. I would love to be able to forecast right now, how many A-grade candidates I will have to work on in January 2011, but unfortunately it doesn't work like that.
But of course the real genius of Atlassian's approach is the free publicity it is generating, and it's innovation will no doubt attract like-minded software engineers directly, rather than necessarily through recruitment firms.
It'll be interesting to learn the outcome of this campaign, and how many were placed by agencies and at what percentage…
Hi Ross, great post – good to see you've presented multiple perspectives on our program.
Jonathan – You might be interested to hear that from our last campaign, Atlassian 32, only 2 of the 69 people we hired were introduced to us by recruitment agencies (from the 60+ who signed up). The rest were all employee referrals, external referrals and direct applicants. It will be interesting to see what the stats are following this campaign.
Sarah,
It sounds to me that Atlassian opens all roles to all recruiters. If all of the 69 roles were actually open to the recruiters and out of all the representations only 2 candidates were hired through recruiters (which is about 3%) I would say it is impressive.
I would also say that for 69 roles you would have to at the very least screen a minimum of 138 resumes coming through recruitment agencies and if out of a minimum of 138 resumes only 2 of them turned out to be a better candidate as compared to advertising/direct referrals I would be concerned about using recruitment agencies for hiring at all.
It takes time to screen that many resumes let alone conducting interviews and if the conversion rate is only 3% the question is does using that avenue pay off even if it was free.