The four objectives, marked

Objective 1, identify the current challenges Accomplished

Met through two independent surveys plus thematic analysis of the free-text answers, checked against the literature.

The evidence shows the failure is recurrent, not exceptional. 92.5 percent of organisers and 82.8 percent of travellers hit it at least sometimes.

One organiser described an event where turnout fell short and, in their own words, resources went to waste and the expense was larger than the profit. That is the precise operational consequence Chapter 1 predicted, said by somebody who lived it.

Objective 2, analyse the statistical relationships Accomplished

The full analytical chain was executed on both samples independently: reliability with purification, descriptives, correlation, and simultaneous regression with collinearity and independence diagnostics.

The result was consistent across two independently recruited groups. All three predictors correlate strongly with the outcome at the bivariate level, in identical rank order. In regression, only centralized information access contributes uniquely.

Objective 3, design and develop the prototype Accomplished

Chapter 5 documents it. The requirement that distinguishes it from every platform reviewed in Chapter 2 is that the unified information layer is a physical property of the schema, not a presentation convention. One inventory table, one availability column, one subscription, one transactional path.

Objective 4, evaluate performance and adoption intent Partially accomplished

Two halves, and they land differently.

Problems encountered, reported without minimising

Four of them. Knowing all four, and being able to state them calmly, is worth more than any defence you could improvise.

1. The achieved sample fell well short of the plan 64 travellers and 40 organisers, against a design of 350 and 50. A combined 104 rather than 400. The organiser sample stays proportionally strong at roughly half the qualifying properties in the collection belt. The traveller sample is both small and narrow: 62 of 64 were domestic travellers and all were aged 18 to 34. The consequence: the traveller model is adequately powered to detect a large effect and underpowered for a small one, which is why the non-significant coefficients are reported as a failure to detect rather than as evidence of absence. The findings generalise to the domestic young-adult segment, not to international visitors.
2. Both reverse-coded items failed psychometrically Items 7 and 20 returned negative corrected item-total correlations on both instruments, depressing initial reliability to 0.159 for the organiser outcome scale. The standard remedy was applied and reported in full, lifting every scale to an acceptable level. The likely cause is a documented method factor, plausibly worsened by respondents answering in a second language. The lesson, stated in the chapter: negatively worded items carry a real risk of measuring reading direction rather than construct position in this setting.
3. The development window was compressed Prototype construction ran against a deadline that left no room for a usability panel, which is why Objective 4 is only partly accomplished. The evaluation that was conducted was prioritised towards the claims that could not be established any other way, concurrency integrity and access control, rather than spread thinly across everything.
4. The outcome could only be measured as perceived, not realised Because the artefact did not exist in the field at survey time, Coordination Effectiveness was captured through behavioural intention as a proxy grounded in the Technology Acceptance Model. A defensible measurement decision, but it is a proxy. The study cannot claim to have measured actual improvement in procurement accuracy or attendance.

Self-reflection

This section is often skimmed by students and read carefully by examiners. Yours has a genuinely strong story in it.

What the project taught, beyond the result

The study required moving between three kinds of work that are usually taught separately: instrument design and psychometrics, statistical inference, and software architecture. The connection between them turned out to be the substance of the project rather than an administrative sequence.

The clearest example: the decision to generalise three resource categories into one table is not a technical preference. It is the regression result expressed in a schema. Being able to trace a line from a survey item, through a beta coefficient, to a database table, to a test case, is a capability no single module teaches.

The three steep learning curves

Business insight

Four points, and they are more specific than "there is a market for this". Each one is constrained by evidence rather than freely chosen.

1. The wedge is the information layer, not the booking engine Since centralized information is what drives adoption intent in both samples, a product entering this market should lead with the unified listing and live availability, not with transaction features. This also lowers the barrier to getting hotels on board: a property can be listed and useful before it ever processes a payment through the platform.
2. The pricing model is constrained by the evidence, not free to choose Organiser item 23, that cost matters more than feature count, returned a mean of 3.600, and cost transparency appeared in 40.0 percent of adoption answers. A per-property subscription is the wrong instrument for a market where 62.5 percent of properties are below mid-size. A commission on completed bookings, or a free listing tier with paid transaction handling, aligns the cost to the operator with the value they actually receive and removes the fixed outlay that would block adoption.
3. Trust has to be manufactured deliberately Demonstrated results and reviews from other operators appeared in 43.3 percent of adoption answers. One organiser said plainly that they would want to run a small or low-risk event with it before trusting it with something large. So: launch on a small number of visible reference properties rather than broad simultaneous acquisition, and build a product affordance that lets an operator trial it on one low-stakes event.
4. The realistic initial market is bounded and reachable Roughly 60 to 80 qualifying properties in the primary collection belt. Small enough to be addressed directly, and dense enough that a traveller-facing listing becomes useful once a modest fraction of them participate, which is the condition any two-sided platform must reach to be viable at all.

Future recommendations

Six of them. Each one answers a limitation, so they are easy to remember if you pair them up.

The concluding statement

Numbers from this chapter

If they ask you

What did you personally learn from this project?

That the most valuable result was one I did not expect. I assumed real-time visibility would dominate, because it is the most visible failure to anyone standing in Mirissa on an event night. The data said otherwise, twice, in two independent samples. Learning to treat that as the finding rather than as a problem with the data was the most useful shift in the project.

The second thing is that instrument design, statistical inference and software architecture turned out to be one continuous piece of work rather than three separate stages. The decision to put tickets, rooms and add-ons in one table is not a technical preference, it is the regression result written as a schema.

Is this commercially viable?

The evidence constrains the answer rather than leaving it open. The entry point is the information layer, not the booking engine, because a property can be listed and useful before it processes a single payment, which lowers the barrier on the supply side.

Pricing is constrained too. A fixed per-property subscription is wrong in a market where 62.5 percent of properties are below mid-size and cost ranked as the deciding adoption factor. Commission on completed bookings, or free listing with paid transaction handling, aligns cost to value received.

And the initial market is bounded and reachable at roughly 60 to 80 qualifying properties in the collection belt, dense enough that the listing becomes useful once a modest fraction participate.

What would you do differently?

Four things, all of them in Chapter 7. Start traveller collection earlier and push much harder on the international segment, which at 3.1 percent is essentially unmeasured. Reword the reverse-coded items or replace them with a different attention check. Build the artefact earlier so a usability panel fits inside the window. And use a longer instrument for the outcome so the three sub-dimensions can be scored separately instead of collapsing into one.

Has this actually improved coordination at a real hotel?

No, and the thesis does not claim it. The evaluation establishes that the artefact conforms to a specification derived from the evidence, and that the concurrency and access guarantees hold. It does not establish realised coordination effectiveness in the field.

Demonstrating that would require deploying across a season with participating properties and measuring procurement variance against actual attendance, the proportion of events running under capacity, and advance turnout prediction accuracy, each against a pre-deployment baseline. That is the first future recommendation in Section 7.5.