The Skill of Manifesting Luck: Aiming at Elusivity with Ruthless Prioritization


Storm” by Craig Armstrong & AR Rahman from the movie “Elizabeth the Golden Age”. 🎵 

“In 1 hour you need to be here.”

It was in that moment I knew that if I did not prioritize the next 30 minutes I would miss the opportunity of a lifetime.

That was Friday.

At 8:30 AM, after going in for my 5th ultrasound, and 7th blood draw in less than 2 months, the technician smiled. “There’s two.”

2 eggs – each 25 mm. I could see them. I grinned. Hard fought – I finally felt like luck was on my side.

I’ve been daily testing using Mira and Clearblue and there was no LH spike. We would have missed it (again). The drugs were working (finally). Being poked weekly paid off in catching ovulation which I can’t easily detect at home and is hard to control or even have with polycystic ovaries. Results don’t happen.

Because of working remotely, I can drive into the doctor at 8 AM whenever they want on whatever day they pick. I can come back after an hour and a half and have productive days, work some west coast if I need to. Work some in India time zone if I need to. This got me to this one try after a year and a half of prioritizing flexibility at work. I will be lucky if even one of those eggs work out – but if both worked out? It would be the forth set of twins in my side of the family generationally. It would also be the second set in the side of my husband’s. I’m trying not to get my hopes up – but even seeing that meant everything to me.

“Should be good to go for tomorrow. We may have to do a trigger. But let’s wait for the bloodwork to come back today.” I breathed out – awesome! We could do the procedure on a Saturday!

…I won’t even have to tell the people I work with…!

Ever looked at a screen and knew the day would not go perfectly? Even when luck is on your side, never underestimate the humor of the universe.

This is for all the women in games who ruthlessly prioritize, say ‘yes’ to the right things for those they care about personally and professionally, and are as flexible as they *rainbows and unicorns* can be.

What Happened Was

My phone blew up at 11 AM, but I missed it. I was working. It kept going off. It was the clinic calling back. “We have to go now. Check your email. Can you drop everything … immediately?” The bloodwork came back. I realized that the slight pains I had been feeling in both my ovaries were signs we actually couldn’t wait.

If you’ve gone through fertility treatment then you may know that a trigger shot is when you get injected with hCG to control ovulation in a window because you are close enough. That was Plan A. Controlling ovulation has eluded me sometimes for over 100 days (Turns out, I do have polycystic ovaries). We didn’t need the trigger shot because the bloodwork showed an LH surge was happening in real-time in the middle of Friday.

I looked at my calendar.

Lol. Shit.

I cancelled everything. To those who said “feel better” because they didn’t know – I feel AMAZING! 🥹
Luck is, finally, on my side.

Here is how I stick to my priorities, trust in what I know works, when I need miracles to happen.

The Exercise: In 30 Minutes, You Lose Today – What did you Prioritize so It Didn’t Matter?

If you know me then you know I don’t like meetings.

love 1:1s and helping people be their best and love what they do, but I don’t like meetings. It’s such an…unproductive word. 1:1s have a goal – me to help you.

If we have a “meeting” then we haven’t prioritized and we don’t know what we need.

I dislike duplication (or triplication) too. I also dislike data that doesn’t tell a story because it doesn’t have value.

Stats for the sake of stats.

Reports for the sake of reporting.

Control for the sake of not giving up ownership.

You need KPIs for the journeys you take when you take them. I personally approach engineering business like this but you do you.

  • Quality: For example, if you care about Kubernetes upgrades and stability as a byproduct of not being on “old shit” then you likely have KPIs for that line of business ownership in the form of how many people are still using deprecated stuff. You track what is deprecated so you can get customers off in time. This can be said of any platform engineering COPE model and isn’t really unique to Kubernetes. K8s, and all the tooling you use for it, just has the bad habit of being known to upgrade at a cadence that is wildly hard to keep up with without a dedicated team.
  • Adoption: If you care about customer adoption, then you have KPIs for onboarding (how long it takes to create the shiny toy for the customer, how many toys they have or asked for), consulting (how much time does it take to teach a customer to use their new shiny toy – how many meetings, how many tickets, support requests), and snowflake-ness projections (who is deviating from the path of onboarding, why, and is that a signal? What do people want you to build because it’s too painful for them to do it?). You may also be tracking “Are people really using the shiny new toy and what for?” (migration metrics, cost metrics, compute split by workloads you vs your customer owns).
  • LiveOps: If you care about uptime and stability then at that moment it’s relevant you report on any increasing anomalies in specific alerting patterns, incidents, or performance lessons. You do not when it is stale. For Kubernetes compute, it is useless to report uptime as a singular metric because it is highly distributed (treat the cluster itself like different lines of business) – many things can go wrong in different parts of the system. If you care about long-term then you track how much your portfolio is deviating from broader industries and the cloud-native ecosystem. You also track applications uptime separately as SLOs (app availability) and, instead, compute uptime based on what is being alerted on in an isolated sub-component when it happens (ex. disk space issues).

Depending on what’s important that week, month, or quarter – swap between these things or something like these things. Don’t report on data that doesn’t matter. The moment it feels useless it probably is. The moment it feels scary – it’s probably valuable. Go where you need to grow. Run at that fear – chances are, scary data is probably telling you something you did not know.

The largest time sync of managers is 1:1s and followed closely by it, customers. This is true whether someone is in a product role, an engineering role, or a production role. Anything else? Is reporting. And reporting? Can be done asynchronously in the direction of who needs it – everyone doesn’t need to know everything.

After delivering awkward news to my husband, I started canceling my calendar for the rest of the day or moving it based on what I see value in. It was easy because I care a lot about the above and thus know “what’s next.”

This exercise was a blessing though – it’s rare I need to be that reactive.
I liked the limitation. It made me test my decisiveness.

20% At the Right Time

It is easy to get in the trap of feeling like because an idea is “good-ish” you have to add it to a backlog – making it hard to be reactive when teams need to be and instead are reactive all the time and prioritizing all the time.

I repeatedly say “No more than 20% of the ideas a business generates, be it coming from the team or coming from customers, is the bar for which a business should chase.” Your number may vary, but have one and tune it based on your years of living in your industry – I’ve seen smaller at scale and provide the 20% as a value to help you measure “what is 20% of ideas?” against time spent.

Teams will generate exponentially more changes, ideas, per engineer added, per customer added, per partner added, per tool used. This means my default answer will almost always be ‘no’ – that’s how I want to operate. I want the ‘yes’ to be hard because I want it to be worth it based on (1) perspectives driven by the scale we are at in a quarter and (2) The faster you let go of ideas the quicker it is to get to the truly amazing ones and also have luck on your side (which you need too). If they are truly that amazing (or needed) they come back through a lesson and can still be handled then almost always.

To break this down as an example – this means in an early stage business if you have 5 people you are directly responsible for, but 100 people you work with (customers, partners, vendors), that means to say ‘yes’ to 10-20% of the asks of the team and 10-20% of customers (1 innovative and new idea from the team, for every 10-20 customer requests) and the broader ecosystem statistically. This is before use cases, the ‘why,’ any KPI is involved, or any opinion from product or engineering to narrow that percentage to something more concrete. That number comes from years of experience seeing volume and that’s often where I’ve landed when I’ve looked back retroactively at what I actually closed and what we actually got done – with 20% being on the exceptionally high side. This is primarily from the perspective of platform and games development in tightly defined business units. It is less in some businesses depending on volume and intake. For example, a VC firm or startup incubator may host a competition, get 200 entries, and only pick 3 to invest in (1.5%). Shark Tank sees 40K businesses try a year but only .4% get admitted, and only .2% get on air. Imagine only doing .2% of your backlog if you treated Jira like an idea factory. Yikes.

If a team, business unit, or startup then has cumulatively 500 people across customers, their own engineers, partners, vendors – I’m still looking for 20% opportunity to invest resourcing in. At scale, you begin to say ‘yes’ to less internally generated ideas and more ‘yes’ to scale-based ideas because you’re now looking for scale based on where your market is. It is a very easy thing to be tempted by “how well a new tool works” or “ease of use” and there can be 10 other things that tick that same bucket. Are we going to say yes to all 10? No. But we will say yes to 2 – how do we know which 2 to say ‘yes’ to and ‘when‘? That’s where your KPIs and quarterly planning come in – and for those you need to measure them over time not in the moment. If you don’t have a KPI for an idea: Figure out which one you need first and then 👏 get 👏 the 👏 quantitative 👏 data. Lean into a quarterly process – never be reactive because it doesn’t let you focus. If you aren’t willing to get the data? The idea isn’t good enough to begin with.

‘Yes’ of value is elusive because rare opportunities are exceptional – the timing matters.

Luck has to be on your side.

Be willing to go on the journey to get enough data to be confident and narrow to the 20% (or less) best ideas based on the quarterly or bi-yearly processes you use to review them.

Be confident enough to be reactive only when it counts (press gas on opportunity instead of reactive from fear) and never anything less.

Header Image by Dustin Humes from Unsplash.