How Consultants Integrate a 48-Hour OEE POC Into a LEAN Engagement
A LEAN engagement at a manufacturing client typically runs eight to twelve weeks. The first two weeks are diagnostic — plant walks, stakeholder interviews, historical data review. Weeks three to ten are improvement sprints. The final week is transition and handover. Within this structure, the single highest-leverage move a consulting team can make in the first two weeks is to get the client real-time OEE data from at least two production lines — not estimated OEE from the client’s ERP, not manually-reconstructed OEE from shift reports, but actual real-time OEE with loss decomposition and root-cause hypotheses.
This is exactly what a 48-hour lightweight OEE POC delivers. This article walks through how consultants at BCG, McKinsey, Bain, Kearney, Roland Berger, Sia Partners, and adjacent firms integrate a TeepTrak 48-hour POC into the diagnostic phase of a LEAN engagement, what to communicate to the client before, during and after, and how to use the output in the steering committee deck at the end of week two.
This article reflects perspectives built during 4 years at BCG working on LEAN programs, operational performance and cost reduction missions across Europe, Asia and the Americas.
When to propose the POC in the engagement lifecycle
The ideal moment is the kickoff meeting at the start of week one, before the team has done any site work. At kickoff, the client expectation is that the first concrete deliverable will come at week two or three — usually a diagnostic presentation with estimated OEE, a loss hypothesis, and a proposed improvement program. If the consulting team can add “and we will have real-time OEE data from two production lines by end of week one” to that expectation, the client’s perception of the engagement tempo shifts immediately.
The POC is a commitment the consulting team should only make if the vendor can deliver reliably. TeepTrak has run 200+ consultant-led POCs and the 48-hour timeline holds if the vendor team is on site with equipment in hand. Adding a travel day or equipment shipment delay pushes the actual delivery to day three or four, which is still fast relative to a traditional measurement infrastructure project but destroys the “48 hours” framing with the client. Plan for physical presence on day zero.
The secondary moment for proposing the POC is mid-engagement, around week four, if the initial diagnostic has surfaced an OEE measurement gap that the client was not aware of. This is common — the client believed their ERP reporting was reliable, the diagnostic showed it was not. The POC then becomes the tool that replaces unreliable ERP-based OEE with reliable real-time OEE going forward. It is a credibility move: the consulting team identified the gap and provided the fix in the same conversation.
Scoping the POC: which lines, how many sensors, what data
The POC should cover two production lines, not one. A single line gives a data point; two lines give a comparison. The comparison is what generates insight for the diagnostic: line A is running at 71% OEE, line B at 58%, and the 13-point gap is concentrated in a specific loss category that the consulting team can then drill into.
Line selection criteria, in order of importance: the lines should be operationally important to the client (not pilot lines or low-volume SKUs), the lines should represent different process types if possible (one packaging line and one assembly line, not two identical stations), and the lines should have obvious OEE improvement potential (the client believes they underperform, even without data).
Sensor coverage per line is typically four to eight sensors depending on station count: current sensors on the main drive motors, photoelectric sensors at product outputs, vibration sensors on critical equipment, and a shared Andon tablet for operator input. The POC kit ships pre-configured; there is no sensor selection decision the consulting team needs to make on site.
Data collected during the 48-hour window: full downtime event stream with timestamps, throughput vs theoretical speed, quality count vs total production, micro-stop frequency and duration distribution, operator-tagged reasons for each event, and environmental signals where relevant (temperature, vibration patterns).
Free Download
Instant download. No email confirmation needed.
The 48-hour timeline, hour by hour
Hours 0-8 on day one are site arrival, safety briefing, sensor installation, network configuration (local WiFi or 4G uplink), and operator tablet setup. The installation is non-invasive: clamp-on current sensors, magnetic-mount vibration sensors, photoelectric sensors mounted to existing frames. No equipment is stopped, no PLCs are touched. The consulting team uses this period for stakeholder interviews on the affected lines — operators, line leads, maintenance staff — which contributes to the diagnostic independently of the POC.
Hours 8-24 are the first production shift with data flowing. At this point the dashboard is live but the data is not yet statistically meaningful. The consulting team should watch the real-time feed to validate the sensors are capturing the events they should — if the line stops and the dashboard shows it stopping within a second, the sensor configuration is correct. If there is a lag or missed events, the vendor engineer adjusts on the spot.
Hours 24-40 are the second production shift. By this point there is enough data for initial pattern recognition. The JEMBA AI layer starts generating hypothesis outputs: “availability losses on line A concentrated in changeover events between SKU-102 and SKU-104, average duration 38 minutes, 3x the target”; “micro-stop frequency on line B is 47 per shift, above the 12-per-shift benchmark, concentrated at station 4.” These are the raw material for the diagnostic deck.
Hours 40-48 are data review, root-cause conversation with client operations leaders, and export of the findings into the consulting team’s deck. The vendor engineer typically leaves site at hour 48; the system continues running and producing data that the consulting team can reference for the rest of the engagement.
What to put in the week-two diagnostic deck
The POC output changes the diagnostic deck in three specific ways. First, the OEE baseline slides move from estimated numbers to measured numbers with a clear methodology footnote — this is the single most credible upgrade a diagnostic can get. Second, the loss decomposition slides have actual timestamps and frequencies rather than hypothesis, which lets the team skip the usual debate with the client about “are these numbers right?” and move directly to “what do we do about them?”
Third, the improvement program proposal gains specificity. Instead of a generic LEAN program proposal, the consulting team can propose targeted interventions: “reduce changeover time from 38 to 20 minutes on line A via SMED workshop in weeks 4-5; this alone should yield 4-5 OEE points”; “investigate station 4 bottleneck on line B via dedicated micro-stop Kaizen in weeks 6-7.” Specificity in the improvement proposal is what separates engagements that close with measurable impact from engagements that close with a strategy deck.
Many consulting teams include the TeepTrak dashboard screenshots directly in the deck — with client permission — because the visual evidence of real-time OEE monitoring reinforces the credibility of the recommendation. The typical format is a two-slide appendix showing the dashboard as-is and a summary of the data quality baseline (event capture rate, sensor uptime, operator entry rate).
Communicating the POC to the client sponsor
Client sponsors — typically plant managers, operations directors, or CIOs — have heard many “48-hour” and “quick-win” pitches before, most of which did not deliver. The credibility bar for this communication is higher than a generic vendor proposal.
Three messages to lead with: the POC does not require any PLC work from the client IT team, the POC is structured to produce decision-quality data within 48 hours, and the POC can be shut down at any point with no residual footprint on the plant’s systems. These three messages address the three concerns that typically kill quick-win proposals: IT workload, time-to-insight uncertainty, and lock-in risk.
The sponsor will often ask what happens at the end of the POC. The honest answer: the consulting team owns the data generated during the POC, the client decides whether to continue with a scaled deployment (typical next step: 4-8 weeks of additional lines) or to stop and take the POC findings forward with other tools. The decision is not pre-loaded. This framing builds trust; the opposite framing (“you’ll love it so much you’ll have to buy the full system”) does not.
Common failure modes and how to avoid them
The POC fails when the consulting team skips the line selection work. A badly chosen line — one where OEE is already high, or where production volume is low during the POC window, or where the line is undergoing maintenance — produces data that does not generate diagnostic insight. The line selection conversation with the client should take at least an hour in kickoff week.
The POC fails when the consulting team does not prep the operators. Operators who see the tablets appear without context will resist. A ten-minute briefing in the pre-shift meeting — “this is a tool to help us understand what is slowing the line down, it is not a performance surveillance tool, no one is getting reviewed on this data” — makes the difference between 90%+ entry rate and 30% entry rate.
The POC fails when the consulting team does not carve out time to review the data. Data that is generated but not reviewed by the consulting team is wasted. Blocking two hours at hour 24 and two hours at hour 40 for the team to sit with the dashboard and the vendor engineer is the minimum. Skipping this ends with a week-two deck that references POC data but has not actually internalized it.
How the POC changes the engagement economics
Consulting engagements with a 48-hour POC at the start consistently produce larger follow-on scope than engagements without. The mechanism is straightforward: the client sees measurable value within the first two weeks, which builds confidence in the consulting team’s technical judgment, which makes the client more receptive to scaled improvement programs, plant rollouts to adjacent sites, and multi-year transformation mandates.
For the consulting firm, the economics are visible at engagement closeout. Engagements with the POC integrated routinely show 1.5-2x the follow-on conversion of engagements without. For the individual consultant, the effect is reputational: clients who saw the consulting team deliver measured OEE improvement within the engagement are materially more likely to be reference-able to the firm’s next prospect, and materially more likely to bring the consulting team back for the next phase of work.
External references: Lean Manufacturing — Wikipedia · OEE — Wikipedia · SMED — Wikipedia
Related reading: Why strategy consultants recommend lightweight OEE over MES · What 4 years at BCG taught me about OEE design · How consultants evaluate MES / OEE vendors for client recommendations
Consulting with a Manufacturing Client?
Our team has supported BCG, McKinsey, Bain, Kearney and Roland Berger engagements across 30+ countries. Free 48h POC on any client site — no client commitment required.
Discuss a Client POC
0 Comments