IGNITE EXTENDED / LESSON 14
Which advantage would actually improve relative profitability?
An advantage scorecard with a pricing-decision memo.
A valuable capability is a competitive advantage when it supports higher profitability relative to relevant competitors over a stated period. Being excellent at an activity is insufficient if competitors are equally capable or customers do not value the difference.
The strategy material asks for the mechanism connecting a resource to profitability. The connection may run through willingness to pay, cost, access or a constraint on imitation. Specify the competitor and the market situation so that the advantage claim can be examined.
An existing firm can enjoy a cost or quality advantage without preventing new firms from entering. A barrier concerns the difficulty of entry or imitation. Conversely, an industry can have high entry barriers while its incumbent firms compete away much of the available profit.
Evaluate the specific source and its expected duration. A patent, customer relationship, scale economy and learning curve operate through different mechanisms. Ask how a challenger could circumvent the source and what investment would be required.
An advantage improves the firm’s performance relative to alternatives. A barrier makes it difficult for another firm to match or neutralize the source of that performance.
A property that can improve profitability relative to relevant competitors.
An advantage specifically against potential entrants; it is one type of advantage.
Illustrative application: a better support process may attract customers today. If rivals can copy it quickly at similar cost, it is a weaker barrier than the current performance suggests. Investigate the knowledge, relationships, activities or accumulated assets required to reproduce it.
Explain both why the advantage creates value and why imitation or substitution may be difficult. The first does not establish the second.
A definition: a competitive advantage is a property of the business or its environment that yields higher profitability, or a higher probability of it, compared with existing or potential competitors. It can show up as higher share, higher prices or lower costs.
The definition has two built-in comparisons: against whom, and in profit terms. 'We have great AI' is not an advantage claim until it names a rival and a profit mechanism.
The page defines a barrier to entry as a property that hinders potential competitors from entering the market, and therefore a special case of competitive advantage.
Barriers are a subset of advantages, scoped to entrants only. A barrier that stops new startups can do nothing against an incumbent that is already inside the market, such as an EHR vendor bundling a feature.
Competitive scope is the set of competitors against which an advantage is effective; many advantages work only against a subset of current rivals or only against future entrants.
Scope is the first question to ask of any moat claim. An integration with an incumbent's API may beat other startups while being irrelevant against the incumbent itself.
The first nine attributes: competitive scope, strength, duration, transferability, replicability, cost to build, time to build, adaptability, and uncertainty about the ability to build.
Replicability and cost to build are the attributes most affected when a capability can be rented: if any rival can buy the same model or service, replication cost falls toward the price of an API key.
The remaining attributes: volatility, dynamics of maintenance costs, long-run sustainability, path to build, and complementarities with other advantages and disadvantages.
Complementarities matter most for small firms: a single weak advantage rarely holds, but several that reinforce each other (integration plus outcome data plus workflow lock-in) can.
The decision point of the case: with negotiations ending, should Mobileye lower its price for a mass-market vehicle program or hold the higher price and risk losing the deal.
The question is not 'which price maximizes this contract' but 'which price builds or erodes future advantage'. Volume can feed learning and installed base, but only if that mechanism is real and rivals cannot match it.
The source evaluates advantages using more than existence. Duration, scope, size, volatility and the path required to build them affect strategic value. An advantage that lasts briefly may still be useful if it creates the resources or learning needed for a stronger position.
Include disadvantages in the same analysis. An early capability can impose a cost or commitment that becomes harmful after the market changes. The relevant comparison is the net effect of the company’s activity system against alternatives, rather than the number of positive items on a list.
A current price affects revenue and also changes adoption, customer learning and competitors’ incentives. The Mobileye discussion uses historical pricing and industry structure to ask how a company can build and sustain its position.
A low price can accelerate adoption or discourage entry, but it can also limit the resources available for development. A high price may capture value while making alternatives more attractive. Explain the dynamic mechanism and test it against the customer’s options. The historical case does not establish a universal pricing rule.
Illustrative diagnosis: evaluate each claim against a named competitor, a mechanism and a time horizon.
A price changes today’s contribution and can also change the path of adoption, learning or capacity use. Strategic analysis must trace the proposed connection.
Price changes margin, adoption and the probability of winning a customer.
A win may create learning, relationships or distribution; a low price may also weaken economics or expectations.
Illustrative application: a lower introductory price may attract useful early customers and produce learning. It may instead attract customers who leave when the price rises or generate demand the service cannot support. Model the immediate cash effect and identify evidence that the proposed future benefit is occurring.
A future-advantage claim needs an operating mechanism and a test. It cannot justify any current loss by itself.
The logic of a business rests on its advantages; for every major decision, ask how it affects those advantages and predict outcomes from them. Some decisions exploit current advantages; others build future ones.
This is the bridge to pricing: a price is also an investment (or a disinvestment) in a future position, and should be evaluated that way.
A timeline: unit pricing near $40-50 under rising competition, the 2016 public split with Tesla, a 2017 roadmap to Level 5 by 2020, Intel's 2017 acquisition, and a 2019 Audi Level 3 note.
Two corrections: Intel paid about $15.3B, and Audi never activated its Level 3 Traffic Jam Pilot for customers, abandoning the plan in 2020. Intel later re-listed Mobileye on Nasdaq in October 2022 at roughly a $17B valuation. The Level 5 forecast shows how far founder roadmaps can run ahead of reality.
A list of potential advantages: production scale and scope economies, and four human-capital advantages (efficiency, alignment, cost, learning curve). The next page adds reputation for threats and punishment, and management sophistication.
Use it as a search list, not a scorecard. Each item still needs scope, mechanism and duration before it counts.
The case compares sensing technologies around 2014: radar sold to automakers for about $100-200 versus $90-150 for a monocular camera system; Google's lidar units cost over $75,000, with cheaper lidar still in development; fusion systems combined sensors.
Mobileye's advantage depended on alternatives' cost curves. A substitute that becomes cheap (as lidar and radar did) can shrink an advantage without anyone copying it, which is exactly the risk falling LLM prices pose to an AI-feature advantage.
Selected reading: PDF pages 2, 3, 6
The selected pages connect a camera-based technical approach with a long commercialization path, capital needs and competing technologies. An advantage must survive both delivery constraints and alternatives.
The founders describe functions they believe can work with a single camera. Their argument combines functionality and cost for selected driving-assistance applications.
The case describes a lengthy path to production adoption. Funding and organizational endurance become part of executing the technical thesis.
The founders seek financing compatible with their expected development path. Later access to capital changes the ability to act during difficult conditions.
The comparison with radar, lidar and sensor fusion shows different performance and cost characteristics in the case’s historical setting.
Selected reading: PDF pages 2, 3, 6
Some comparisons and forecasts come from the founders. Distinguish their assessment from an independently established performance comparison.
Prices, regulatory expectations and driving forecasts belong to the case period. The transferable question is how to analyze alternatives and adoption, rather than reuse those numbers today.
A useful feature or lower cost establishes only part of a moat argument. Ask how rivals could match it and which accumulated capabilities would be difficult to reproduce.
The historical case asks whether to lower a price for a mass-market opportunity or risk losing the deal. The decision involves near-term economics as well as the potential effect on adoption, relationships and future competitive position.
Which advantages would each pricing path build or weaken?
Epoch AI estimated in March 2025 that the price of reaching a fixed LLM performance level fell between 9x and 900x per year depending on the benchmark, about 40x per year for GPT-4-level science questions. Its authors caution that the fastest declines may not persist. An advantage that rests on the model itself erodes on that schedule.
At its August 2025 user meeting Epic introduced AI agents, including Emmie, a patient-facing assistant in MyChart. In March 2026 Epic reported that patients at Ochsner Health had rescheduled more than 14,900 appointments with Emmie. This is the ClinicOS bundle threat in real life: the incumbent owns the patient channel and the schedule.
In September 2025 HHS announced stepped-up enforcement of the 21st Century Cures Act information-blocking rules; certified health IT developers face penalties of up to $1 million per violation. Whether a specific API restriction counts as information blocking depends on the facts and current rules and needs legal review.
Sources: Epoch AI: LLM inference prices have fallen rapidly but unequally across tasks · Healthcare IT News: Epic unveils AI agents, showcases new foundational models · Fierce Healthcare: HIMSS26, Epic expands AI roadmap · HIPAA Journal: HHS announces crackdown on information blocking · Ropes & Gray: OIG announces penalties of up to $1 million per information blocking violation (2023)
Fictional company claims: a patented component, two years of field data, and a distributor relationship. A large buyer asks for a price reduction. Competitors offer a cheaper but less proven alternative.
For each claim, specify scope, mechanism, likely duration and a threat. Explain how the pricing decision might change one advantage.
Relay ends FY1 with 60 clinics at $399 per location per month. ClinicOS, the EHR used by about 35% of independent PT clinics, can bundle a basic waitlist feature free, and Relay depends on ClinicOS's API. Any rival can now rent a capable LLM to text patients, and the planned Relay Scheduler would run on such an API. Relay's candidate advantages: EHR write-back integrations, acceptance data from thousands of slot offers, clinic workflow embedding, local patient waitlists, and its HIPAA program. A clinic owner asks for a price cut to $299 to match the free bundle.
Build an advantage scorecard: for each candidate, state scope (against which rival), mechanism (willingness to pay, cost, access), duration and the main threat. Then write a one-paragraph pricing memo answering the $299 request.
A capability rented from an LLM vendor is available to every rival at the same price, and that price keeps falling. Ask what remains scarce if a competitor uses the identical model tomorrow.
Record counts are not an advantage. Show that the data improves an outcome customers pay for, and check whether the rival you fear, here ClinicOS, holds more of the same data.
An integration that takes startups a year to copy is no advantage against the vendor that owns the API. Score every claim against a named competitor and redo the scorecard for each one.
Relay's advantage scorecard (five candidates with scope, mechanism, duration and threat) and a pricing memo holding $399 against ClinicOS's free bundle on the basis of about $48,000 of incremental value per clinic.
Next lesson: Strategic change, pivots & organizational alignment
Check your understanding
Question 1
ClinicOS announces a free basic waitlist inside its EHR. Which Relay asset is most likely to remain an advantage against ClinicOS specifically, rather than only against other startups?
ClinicOS cannot credibly serve rival EHRs, so multi-EHR neutrality has scope against it. The ClinicOS integration is tempting but it is an advantage only against startups lacking it, and ClinicOS controls its terms. AI texting and SMS are replicable by anyone.
Question 2
A clinic has 42 empty slots a week. ClinicOS's free bundle fills 20% of them; Relay fills 40%. Average reimbursement is $110. What is Relay's incremental annual value over the free alternative?
Incremental fills are 42 x (40% - 20%) = 8.4 visits a week; 8.4 x $110 x 52 = $48,048. The $96,000 figure compares Relay with doing nothing and the $66,000 figure with the phone list; once the bundle exists, the bundle is the customer's alternative.
Question 3
A market has high entry barriers: a dominant platform, regulatory approvals and long sales cycles. What follows for the profitability of the firms already inside it?
A barrier is an advantage scoped to potential entrants. Existing rivals can still undercut each other on price, so high barriers and low incumbent profits can coexist. Treating a barrier as a profit guarantee confuses scope.
Question 4
Relay claims a 'data moat' from roughly 130,000 empty slots a year across its clinics (60 x 42 x 52, illustrative). Which evidence would best support the claim?
Data is an advantage only through a mechanism: better predictions that change fills and therefore willingness to pay. Volume alone is not a mechanism. ClinicOS's current inaction says nothing about how hard imitation would be.
Question 5
Mobileye must choose between cutting its price to win a mass-market vehicle program and holding price. Under which condition does the price cut build advantage rather than just give up margin?
A low price is an investment only if volume feeds a real mechanism, such as learning, scale or installed base, that competitors cannot replicate in time. Winning a big deal is not by itself an advantage.
Question 6
Relay's planned AI scheduler runs on a third-party LLM API. Which advantage attribute is weakest for the AI conversation feature itself?
Any competitor, including ClinicOS, can rent the same model, so replication cost is low. Time to build is short too, but that is a consequence of replicability. The feature may still complement stronger advantages such as integrations.
Workbook
Module output: A business-model map, market/game models and an advantage-and-pivot memo.
Sources
Original pages and full documents are in the Library edition.