Pivot or Persevere: Count the Weak Traction Readings

A startup should iterate when two of its four traction readings are confirmed weak and open a pivot review when three or four are, according to The Startup Pivot Playbook by Nitin Kumar. The four readings are ICP share of new bookings, net revenue retention, CAC payback and burn multiple, and a planned change that scores 0 to 4 out of 18 on the pivot score counts as an adjustment, the book’s term for an iteration.

The book reads the count each quarter against fixed lines, then sizes whatever change the count calls for with the pivot score. A board that sees both numbers argues about the evidence in front of it, and the founder walks in with the same figures the investors will ask for.

The traction readings and their red lines

Table 1 · Traction readings and red lines from Chapter 3 of The Startup Pivot Playbook

Working defaults for B2B software from seed to growth stage. A reading counts as confirmed weak when it meets the persistence conditions in the book’s traction scorecard, which look at prior quarters as well as the current one.

The decision rule

One weak reading calls for a closer look next quarter, since a single number past its line usually has a local cause: one channel that stopped converting, one cohort that churned, one pricing tier set too low. Two confirmed weak readings call for an adjustment aimed at those two parts of the business. Three or four confirmed weak readings open a pivot review, and the book designs a pivot only after that review and a runway gate.

The count works because the readings fail together for a reason. When the ideal customer share, retention and payback all weaken at once, the cause usually sits in the target market or the product, which sit upstream of every metric, and an iteration on the channel or the price cannot reach them. When one or two readings slip, the rest of the business is still telling the company its market and product hold.

Adjustments in the database

All 4 adjustments in Nitin Kumar’s database of 75 scored pivots succeeded. Arm & Hammer scored 2 in 1972 when it sold the same baking soda as a refrigerator deodorizer, and Old Spice scored 3 in 2010 when it aimed its body wash at the women who bought it for their partners. PTC scored 4 for the move to subscription licensing it began in 2016, and Dunkin’ scored 4 in 2018 by leading with drinks over donuts. Each kept the product, the team and the cost base and changed the use, the message or the price.

The line between an adjustment and a pivot is one point on the score. Autodesk made the same kind of move in 2016 and scored 5, a hard pivot. It stopped selling new perpetual licenses for most individual products after January 31, 2016 and for most suites after July 31, 2016, worldwide. PTC phased its change, kept perpetual licenses on sale in some regions and made subscription the only option for new licenses in the Americas and Western Europe from January 2018. Both succeeded, and the score told each board how far its change reached before it began.

Table 2 · Scored cases near the line between an adjustment and a pivot

Method 2026.10, dataset of 75. Every case scoring 0 to 5 in the database. Year is the year the move began.

Small moves against large threats

The two failures at a score of 5 made small moves against problems that needed large ones. Blockbuster added mail and online rental in 2004 while Netflix was building the habit that would replace stores, and the stores, late fees and brand stayed as they were. Tribune renamed itself tronc in 2016 with the same newspapers behind the new name and reversed it in 2018. In the book’s terms, both treated a threat to the core of the business as an adjustment.

The traction count guards against that mistake from the other side. A company with three or four confirmed weak readings that answers with a price change or a new tagline is iterating on a problem the readings have already placed upstream, and the two failures at a score of 5 show where that answer tends to lead.

A worked example

Take an illustrative B2B software startup with these readings for last quarter: ICP share of new bookings 42%, net revenue retention 96%, CAC payback 27 months and burn multiple 1.8x. Three of the four sit past their red lines, with only the burn multiple inside. If the prior quarter shows the same three weak, the scorecard confirms them and the rule opens a pivot review.

The runway gate comes next. Say the company holds $3 million and burns $150,000 a month, which gives 20 months of cash. The book’s runway rule asks for enough cash to reach the first proof point of the new direction plus six months. If the first proof point is a paying customer in a new segment 12 months out, the rule asks for 18 months, and the company has 20, so it can design the pivot. If the first proof point sits 16 months out, the rule asks for 22 months, and the company needs a bridge or a nearer proof point before it commits.

Change one number and the answer changes. With retention at 104%, only ICP share and payback sit past the line, two weak readings, and the rule calls for an adjustment: tighten the ideal customer profile with Chapter 10 and fix the channel with Chapter 11, while the product and the team stay. The free traction test runs this count on your own numbers in the browser, with no checkout and no email.

Pivot or persevere, with numbers

Eric Ries named the pivot-or-persevere meeting in The Lean Startup in 2011. The Startup Pivot Playbook adds two numbers to that meeting: the count of confirmed weak readings, which decides between adjusting and opening a pivot review, and the pivot score of the planned move, which tells the board how large the change will be. The Startup Pivot Playbook vs The Lean Startup compares the two books, when to pivot a startup covers the signals in more depth, and the startup pivot guide sets out the order of moves.

Questions

Should a startup pivot or persevere?

The Startup Pivot Playbook by Nitin Kumar answers with a count of four traction readings: persevere with an adjustment when two are confirmed weak, and open a pivot review when three or four are. A planned change that scores 0 to 4 out of 18 on the pivot score is an adjustment.

When should a startup iterate instead of pivot?

The Startup Pivot Playbook by Nitin Kumar says to adjust, its term for iterating, when two of four traction readings are confirmed weak, and to open a pivot review when three or four are. The readings are ICP share, net revenue retention, CAC payback and burn multiple.

What is the difference between a pivot and an iteration?

In Nitin Kumar’s pivot score, a change that scores 0 to 4 out of 18 is an adjustment, the book’s word for an iteration, and 5 to 13 is a pivot. All 4 adjustments in his database of 75 scored pivots succeeded.

Is a pricing change a pivot?

Usually not, in Nitin Kumar’s database. PTC’s phased move to subscription licensing scored 4, an adjustment, while Autodesk’s worldwide end of new perpetual licenses in 2016 scored 5, a hard pivot.

What are the signs a startup needs to pivot?

Nitin Kumar’s Startup Pivot Playbook reads four numbers: ICP share of new bookings below 50%, net revenue retention below 100%, CAC payback above 24 months and burn multiple above 2.5x. Three or four confirmed readings past those lines open a pivot review.