A manufacturing client was about to commit $ $8 million to a new production line. The investment committee had already approved the budget. The supplier had the timeline ready. The signature was missing.<\/p>\n\n\n\n
Before that moment arrived, we asked ourselves a question that should be standard in any CAPEX approval and almost never is: what happens if we simulate this decision before committing capital?<\/p>\n\n\n\n
The response changed the outcome. Not because the line design was flawed \u2014 the engineers had done their job well \u2014 but because no one had modeled what could go wrong in the actual operation before approving the expenditure. The committee was about to make an $8 million decision based on the supplier\u2019s certainty, not on a model of their own decision-making.<\/p>\n\n\n\n
This is what I see time and again in capital investments: the technical rigor of the design has nothing to do with the rigor of the decision to invest. The engineering is audited. The price is negotiated. The timeline is reviewed. But almost no one sits down to model what happens if the baseline scenario doesn\u2019t materialize.<\/p>\n\n\n\n
This is not an issue exclusive to manufacturing. It is a problem of how investment committees generally approve business cases: the majority are approved based on the investment thesis of the seller or the proposer, not on an independently validated own investment thesis.<\/p>\n\n\n\n
The difference between simulating and guessing is not semantic. Guessing is assuming that the baseline scenario will be fulfilled because the document says it will be fulfilled. Simulating is building a decision model that tells you, with data, how sensitive your return to the assumptions you are taking for granted is.<\/p>\n\n\n\n
In this case, before the customer committed capital, we built a digital twin of the proposed production line. It wasn\u2019t an exercise in detailed engineering\u2014it was a decision model: What happens to the return if the start-up takes longer than planned? What happens if the quality rejection rate in the first few weeks exceeds what the supplier\u2019s baseline assumption assumes? What happens if the supply of critical inputs has the variability we\u2019ve already seen in comparable lines?<\/p>\n\n\n\n
Each of these scenarios, tested in the digital twin, cost zero additional pesos. Testing them in the real plant, after committing the $8 million, would have cost months of unplanned inefficiency at best, or a process redesign after implementation, at worst.<\/p>\n\n\n\n
The finding was not a design error. It was a risk in the startup sequence that the vendor\u2019s baseline did not account for\u2014the kind of risk that only appears when someone takes the time to model the adverse scenario, and not just the optimistic one. (The exact numbers of this case are customer information; what we share is the process and the type of finding, which is replicable in any capital decision of this size.)<\/p>\n\n\n\n
Here is the part that almost no one discusses when talking about digital twins: the model is as reliable as the data that feeds it. 70% of analytics projects fail due to data quality issues, not model issues. You can have the best simulation algorithm in the market and still make a bad decision if the input data\u2014historical cycle times, failure rates, variability of inputs\u2014is not reliable.<\/p>\n\n\n\n
Therefore, before simulating, we did the Data Quality work: cleaning, validating, and reconciling the client\u2019s operational data sources \u2014 maintenance data, quality data, and supply chain data \u2014 so that the decision model would not inherit the same biases and gaps that already existed in the source systems. A digital twin built on dirty data does not reduce the risk of the decision. It disguises it with a more sophisticated interface.<\/p>\n\n\n\n
This is the discipline we apply in any digital twins or decision-making model initiative: first the quality of the data, then the model. In that order. Inverting the order is the reason why many Industry 4.0 initiatives produce attractive dashboards and equally risky decisions as before.<\/p>\n\n\n\n
The first objection we hear at discovery is time: the committee has already approved, the provider already has a start date, stopping the process to simulate feels like unnecessary friction. In this case, the data quality diagnosis and the complete decision model were resolved in weeks, not months \u2014 considerably less time than it would have taken the customer to discover the start-up risk once it was operational.<\/p>\n\n\n\n
The second objection is to trust in existing data: \u00abwe already have the plant\u2019s data, we already know how it behaves.\u00bb It is true that, almost always, the data exist. What is almost never true is that they are clean, reconciled between systems, and ready to feed a model of decision-making without bias. Confirming that\u2014or correcting it\u2014is the first step, not an optional step.<\/p>\n\n\n\n
The client did not cancel the project. It approved it, but with a tight start-up plan to absorb the risk that the digital twin had identified \u2014 and with a committee that, for the first time in that process, saw the full sensitivity of the return before signing, not just the vendor\u2019s optimistic scenario.<\/p>\n\n\n\n
That is the real change that the simulation produced: not necessarily a different number in the business case, but a decision made with an open eye to the risk, rather than with blind trust in the baseline case of the seller.<\/p>\n\n\n\n
If you have a large capital investment in your committee\u2019s decision pipeline, the question you should be asking is not \u00abWhat guarantees the provider?\u00bb It is: \u00abWhat own decision model do we have, built on reliable data, before committing capital?\u00bb<\/p>\n\n\n\n
First, data quality diagnosis: what sources exist, how reliable they are, what gaps need to be closed before any simulation makes sense. Second, decision model: not necessarily a digital twin of detailed engineering, but a model that answers the specific questions the committee needs to approve rigorously. Third, business case with scenarios: not a single number, but a range, with the contingency budget sized by data. Fourth, the decision: with the committee looking at not only the vendor\u2019s baseline case, but the full sensitivity of the return.<\/p>\n\n\n\n
The same pattern repeats in any sector where a capital or business decision is approved before it has been tested. In retail, we see it in sell-in and sell-out decisions that are approved based on the historical performance of a channel without modeling the variability of actual demand. In banking, in origination decisions that rely on a single client when origin systems still have duplicates and inconsistent data across channels. In textiles, in seasonal production decisions that are committed before simulating the sensitivity of margin to the variability of imported inputs.<\/p>\n\n\n\n
In all cases, the error is not sector-specific. It is a sequence error: it is decided first and validated later, when it should be the other way around. And in all cases, the same principle applies\u2014the decision model is as good as the data governance that supports it. That is why the practice of Data Quality, Single Customer, and MDM is not an isolated IT project. It is the entry condition for any simulation, in any industry, to be more than just a visualization exercise with a veneer of rigor.<\/p>\n\n\n\n
Most investment committees require a business case. Very few require a sensitivity analysis of the return to critical assumptions before approving. That distinction\u2014request versus demand\u2014is what separates organizations that make disciplined capital decisions from those that make them with well-presented optimism.<\/p>\n\n\n\n
Demanding sensitivity does not mean distrusting the team that brings the proposal. It means recognizing that no team, no matter how good, has the incentive to actively model the scenario in which their own project does not work as planned. That exercise needs to come from outside the team proposing it, with the same rigor with which any other financial assumption of this size is audited.<\/p>\n\n\n\n
The next time your committee is about to approve a large capital business case, before signing, ask the same question we asked this client: what if we simulate this first?<\/p>\n\n\n\n
Request a diagnostic session with our specialists to assess the feasibility of your investment projects<\/p>","protected":false},"excerpt":{"rendered":"
Un cliente de manufactura estaba a una firma de comprometer $8 millones de d\u00f3lares en una l\u00ednea de producci\u00f3n nueva. El comit\u00e9 de inversi\u00f3n ya hab\u00eda aprobado el presupuesto. El proveedor ten\u00eda el cronograma listo. Faltaba la firma. Antes de que llegara ese momento, nos hicimos una pregunta que deber\u00eda ser est\u00e1ndar en cualquier aprobaci\u00f3n […]<\/p>\n","protected":false},"author":3,"featured_media":344567,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_et_pb_use_builder":"off","_et_pb_old_content":"","_et_gb_content_width":"","content-type":"","footnotes":""},"categories":[1],"tags":[],"class_list":["post-344566","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-blog"],"_links":{"self":[{"href":"https:\/\/scanda.com.mx\/en\/wp-json\/wp\/v2\/posts\/344566","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/scanda.com.mx\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/scanda.com.mx\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/scanda.com.mx\/en\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/scanda.com.mx\/en\/wp-json\/wp\/v2\/comments?post=344566"}],"version-history":[{"count":1,"href":"https:\/\/scanda.com.mx\/en\/wp-json\/wp\/v2\/posts\/344566\/revisions"}],"predecessor-version":[{"id":344568,"href":"https:\/\/scanda.com.mx\/en\/wp-json\/wp\/v2\/posts\/344566\/revisions\/344568"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/scanda.com.mx\/en\/wp-json\/wp\/v2\/media\/344567"}],"wp:attachment":[{"href":"https:\/\/scanda.com.mx\/en\/wp-json\/wp\/v2\/media?parent=344566"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/scanda.com.mx\/en\/wp-json\/wp\/v2\/categories?post=344566"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/scanda.com.mx\/en\/wp-json\/wp\/v2\/tags?post=344566"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}