The 47,400 consulting jobs that went vertical on slide 4 and flat on slide 17
Showing the misleading chart
A partners’ deck we drew uses the same BLS series twice: on slide 4, in a tall narrow box, consulting employment has “gone vertical”, so hire now; on slide 17, in a wide short strip, it is “flatlining”, so freeze raises. Same 34 months, same 1,850–1,910 axis, same gridlines, same source line. The latest rise draws at 73° on one slide and 9° on the other. In jobs it is +47,400, or 2.6% in fourteen months: faster than the economy, slower than any fourteen months of 2015–2020.
01The claim
Slide 4: consulting headcount has gone vertical, so hire now before talent costs more. Slide 17 of the same deck: industry headcount is flatlining, so there is no talent war and raises can be capped.
02The trick
Both slides plot the same 34 monthly BLS values, January 2024 to October 2026, on the same 1,850-to-1,910 thousand axis, with the same three gridlines, the same line colour and the same source line. What changes is the shape of the plot box: 150 px wide by 270 px tall on slide 4, 680 px wide by 60 px tall on slide 17. A slope is set by two exchange rates, pixels per 1,000 jobs up the side over pixels per month along the bottom, and no chart prints either one. Slide 4 gives a month 4.55 px and a thousand jobs 4.50 px. Slide 17 gives a month 20.61 px and a thousand jobs 1.00 px, so slide 4 runs the vertical rate 20.4 times harder. The rise from August 2025 (1,858,300) to October 2026 (1,905,700) therefore draws at 73.4° on the “gone vertical” slide and 9.3° on the “flatlining” one. Both headlines are claims about an angle, and the box sets the angle. The figures in the bullets are right, which is why the deck gets through review: 47,400 jobs since August 2025 is correct, and so is +2.4% in 33 months. Each slide picks the box and the figures that suit its recommendation. The one bullet that goes beyond the data is “still steepening”: September 2026 fell by 3,600. (We drew this deck and its firm as a demonstration; the employment figures are BLS’s.)
03The fix
The redraw keeps the series and the axis, and prints what the slides left out. It banks the box so the comparison the deck argues about, August 2025 to October 2026, sits at 45°, says so in the subtitle, and lists all three boxes with their pixels per month, pixels per 1,000 jobs and the angle each one draws. The rule for a deck follows from that: choose one box for a series and keep it on every slide that reuses it, the way small multiples keep one axis. Then the redraw takes the weight off the angle. The levels go on the chart: 21 of the 22 months from January 2024 to October 2025 sat between 1,853,500 and 1,866,600, apart from a one-month spike to 1,880,000 in December 2024, before the climb to 1,905,700. The axis still starts at 1,850 to match the slides, so the height still flatters the rise, and a second panel gives its size in percent over the same fourteen-month span. Consulting employment grew 2.6%, about five times the 0.5% of all US jobs. But every fourteen-month span from January 2015 to February 2020 grew faster, between 2.9% and 7.4% with a median of 4.0%, and the rebound from January 2021 to March 2022 managed 11.9%. So neither slide has it right: the industry is growing faster than the economy and slower than its own pre-pandemic pace. The endpoints also help slide 4. August 2025 is a local low, and September and October 2026 are preliminary estimates, with October alone supplying 17,300 of the 47,400. Measured August 2025 to August 2026, the latest month no longer preliminary, the year’s growth is 1.8%.