Latest / Modularity Podcast / Why Full Calendars Break Small Businesses

Transcript

Why Full Calendars Break Small Businesses · Modularity Podcast

• 0:00
  Your calendar's booked solid for the next six weeks.
• 0:03
  Every crew is dispatched, every hour is sold.
• 0:06
  Is that a great business, or a business about to fall over?
• 0:09
  Um, I mean...
• 0:12
  the first one?
• 0:13
  That's the dream, right?
• 0:14
  Full calendar, full revenue.
• 0:17
  That's the instinct.
• 0:18
  And I think it's, it's, it's half right in a way that hurts you.
• 0:23
  Here's the mystery.
• 0:24
  Why are the busiest small businesses the ones that run weeks behind?
• 0:29
  Plumbers, landscapers, agencies.
• 0:32
  Booked flat out, always late, team fried.
• 0:35
  I've seen it in dev shops.
• 0:36
  Every hour of the sprint accounted for, and then one bug blows up the whole thing.
• 0:41
  Right.
• 0:42
  And there's math on it.
• 0:43
  A British mathematician, Sir John Kingman, published a queuing formula in 1961.
• 0:49
  According to the write up on All About Lean, it estimates how long work waits,
• 0:53
  based on two things.
• 0:54
  How busy the system is, and how much variation there is.
• 0:57
  Variation meaning...
• 1:00
  Jobs don't show up evenly, and they don't take the same time.
• 1:03
  One's a ten minute fix, the next one you open the wall and find rot.
• 1:07
  Okay, so where's the surprise?
• 1:10
  The busy part of the formula is utilization divided by one minus utilization.
• 1:16
  At fifty percent busy, that number is one.
• 1:20
  At eighty percent, it's four.
• 1:23
  At ninety, nine.
• 1:26
  At ninety five, nineteen.
• 1:29
  Wait, wait.
• 1:30
  So ten more points, eighty to ninety, and the wait more than doubles?
• 1:35
  More than doubles.
• 1:36
  And the source says the closer you get to a hundred percent utilization,
• 1:40
  the closer you get to an infinite queue length.
• 1:44
  Infinite.
• 1:44
  Come on.
• 1:45
  Nobody's waiting forever.
• 1:46
  Fair.
• 1:47
  It's an approximation, and the article says so itself.
• 1:50
  In real life people cancel, you work weekends.
• 1:53
  But the shape is right.
• 1:55
  A ScanQueue explainer on queuing theory says above about eighty five percent,
• 1:59
  and wait times rise sharply.
• 2:01
  So it's not a ramp.
• 2:02
  It's a cliff.
• 2:03
  Flat, flat, flat, then basically vertical.
• 2:06
  Give me the ugly version.
• 2:08
  What does the cliff look like on a Tuesday?
• 2:10
  A tech gets stuck in traffic.
• 2:12
  Or a part shows up two hours late.
• 2:15
  Or a routine appointment turns into, oh, there's damage behind that.
• 2:21
  With slack, you absorb it.
• 2:24
  With zero slack, those two hours have nowhere to go, so they push every job after
• 2:30
  them.
• 2:30
  Tomorrow's already full, so the delay carries over.
• 2:34
  Next thing you know, one bad morning is a three week slip.
• 2:38
  It's a server pinned at full CPU.
• 2:40
  One little spike and every request behind it just stacks up.
• 2:44
  Yeah.
• 2:45
  Same shape.
• 2:47
  But the hard part isn't the math.
• 2:49
  It's the feeling.
• 2:51
  You look at an empty Tuesday afternoon and your stomach turns.
• 2:55
  I'd fill it.
• 2:55
  I, I, I would absolutely fill it.
• 2:57
  Everybody does.
• 2:58
  Employee brain says idle time is waste.
• 3:02
  CEO brain has to see that empty block as doing a job.
• 3:06
  Unlearning that reflex might be the hardest shift for a founder.
• 3:10
  Okay, so what job is it doing?
• 3:13
  Shock absorber.
• 3:15
  Good service businesses design the schedule around eighty to eighty five percent,
• 3:19
  on purpose.
• 3:20
  The leftover room handles emergencies, lets you grab a same day job at a premium,
• 3:25
  and cuts the overtime you'd otherwise pay for catching up.
• 3:28
  But I'm still paying people for that empty fifteen percent.
• 3:31
  Somebody's gotta fund it.
• 3:32
  Pricing funds it.
• 3:34
  Two moves.
• 3:35
  Raise prices ten to fifteen percent, and let go of the bottom ten percent of
• 3:39
  clients.
• 3:40
  The high maintenance, low margin ones.
• 3:42
  Hold on.
• 3:43
  Raise prices and lose clients?
• 3:46
  That sounds like shrinking.
• 3:47
  Rough math, and I'm assuming those clients are average size.
• 3:51
  Keep ninety percent of clients, charge them ten percent more,
• 3:55
  and you're at about ninety nine percent of the old revenue.
• 3:58
  Charge fifteen more, about a hundred and three and a half.
• 4:02
  Revenue's roughly flat.
• 4:04
  And you've got ten percent of your capacity back.
• 4:06
  Right.
• 4:06
  Same money, less load.
• 4:08
  And the clients you shed are usually the ones who eat the most hours,
• 4:11
  so it's probably better than that.
• 4:13
  Okay.
• 4:14
  What do people do this week?
• 4:15
  Two things.
• 4:16
  One, audit capacity against what you've actually promised.
• 4:19
  Person by person, crew by crew.
• 4:21
  You'll probably find a hidden ninety percent plus spot, someone everyone leans on.
• 4:26
  Always that one person, yeah.
• 4:28
  Two, put a recurring half day buffer on the calendar.
• 4:31
  Protected.
• 4:32
  Not for new work.
• 4:34
  For friction.
• 4:35
  And if nothing goes wrong that week?
• 4:37
  Then you catch up, or you sell it as a same day slot.
• 4:40
  Either way it's paid for itself.
• 4:42
  So the empty block isn't, um, wasted overhead.
• 4:45
  It's what lets you charge more and show up on time.
• 4:48
  That's it.
• 4:49
  Buffer is pricing leverage and customer trust.
• 4:53
  If you want help engineering systems that hold up, modularity.us.
• 4:58
  So the full calendar's the one that falls over.
• 5:00
  Good to know.