sualiha

Every hardware startup eventually reaches the same fork in the road: does the product need a full-time mechanical engineer on the payroll, or does it need outsourced product engineering support instead? There is a third option too, one that most founders do not consider until they have already made an expensive mistake with the first two: a hybrid model that combines internal ownership with external engineering capacity. The short answer is this. Hire in-house when mechanical engineering is a continuous, long-term part of your operation. Outsource when your engineering needs are project-based, specialized, or time-sensitive. Use a hybrid model when you want to keep product ownership internal while accessing external expertise for the parts of the build you cannot yet staff yourself. Which of these fits your startup depends on your product stage, your engineering workload, the expertise you are missing, your timeline, and your runway. This guide walks through each of those variables so you can make the decision with your eyes open, rather than defaulting to whichever option a recruiter or an agency pitched you first. Not sure whether to hire in-house or outsource your mechanical engineering? Get expert support for your hardware product without committing to a full-time engineering team. See how the right engineering model can fit your product stage, workload, and timeline. Engon’s mechanical product engineering support for startups. EXPLORE MECHANICAL ENGINEERING SERVICES → Why This Decision Is Difficult for Hardware Startups Software startups can scale engineering headcount up and down with relative ease. Hardware startups cannot. A hardware product has to move through physical design, prototyping, testing, and manufacturing, and the skills required at each stage are not identical. The mechanical design work needed to build a first prototype looks very different from the design-for-manufacturing work needed to build the thousandth unit. This creates a few structural problems that software companies rarely face: Engineering requirements shift significantly as a product moves from concept to prototype to production. Different stages need different specialists, not just more of the same specialist. Workload is rarely constant. A mechanical engineer might be essential for four months and underused for the next four. Manufacturing introduces requirements, tolerances, materials, and supplier constraints that a generalist engineer may not have deep experience with. Runway is limited, and hiring takes time you may not have. One engineer, however capable, cannot cover mechanical design, industrial design, electronics, and manufacturing engineering simultaneously. The key message is simple: a startup does not need to build every engineering capability internally from day one. The decision is not “hire or don’t hire.” It is “What is the right way to access the capability we need, right now, at this stage of the product?” What Does a Mechanical Engineer Actually Do for a Hardware Startup? Before deciding how to access mechanical engineering capacity, it helps to be precise about what the role actually covers. A mechanical engineer working on a hardware product is typically responsible for: Mechanical architecture and system-level layout Mechanical design and 3D CAD Component and assembly design Enclosure design FEA Simulation and Design Validation Material selection Tolerance analysis Tool Design and Engineering Support Engineering drawings and documentation Prototyping and design validation Design for manufacturing (DFM) Communication with manufacturing partners Managing design revisions as feedback comes in from testing or suppliers Here is the distinction that matters most for this decision: hiring one mechanical engineer does not automatically give you a complete product engineering capability. A physical product often needs industrial design, electronics, embedded systems, simulation or CAE, testing, and manufacturing engineering alongside the core mechanical work. A single hire covers one piece of that puzzle. That gap is exactly why outsourcing or a hybrid model becomes relevant for many startups, even ones that eventually plan to hire internally. When Should a Hardware Startup Hire an In-House Mechanical Engineer? Hiring makes the most sense when mechanical engineering is going to be a permanent, ongoing part of how your company operates, not just a one-time push to get a product built. Mechanical Engineering Is a Core Long-Term Capability If mechanical design and iteration will remain a continuous part of your business for years, not months, an internal hire starts to make financial and strategic sense. You are not paying for a project. You are building a competency the company will rely on indefinitely. You Have a Consistent Engineering Workload A consistent engineering workload can make either a full-time hire or an ongoing outsourcing partnership a practical option. A steady backlog of design changes, new SKUs, or product variants may justify an in-house engineer, particularly when deep product knowledge and daily internal collaboration are important. However, outsourcing can also be economically attractive when a company needs ongoing engineering capacity without building or expanding its permanent team. An engineering partner can provide continued support, access to specialized expertise, and flexibility as requirements evolve. The right choice depends on the total cost of each model, the capabilities required, the level of internal collaboration needed, and the long-term engineering workload. You Need Deep Internal Product Knowledge An internal engineer accumulates institutional knowledge that is hard to replicate externally: why a decision was made eighteen months ago, how a specific customer complaint shaped a tolerance, what failed in testing and why. That accumulated context has real value for products with long lifecycles and frequent iteration. You Plan to Build a Larger Engineering Organization If your roadmap includes building an internal engineering team of five, ten, or twenty people, an early mechanical engineering hire can become the foundation of that team, not just a solo contributor. Product Development Requires Daily Internal Collaboration Some products need engineering decisions made in real time, side by side with electronics, firmware, manufacturing, and operations teams. If your development cadence genuinely requires that level of daily, in-person coordination, an internal hire removes friction that a remote or external arrangement would introduce. Engineering Scenario: Multiple Products, Continuous Engineering Work A hardware company has several products in development and production, with regular design changes and supplier coordination. Engineering Insight:

Choosing a mechanical engineering partner is not the same as choosing a CAD vendor. A CAD vendor draws what you tell them to draw. An engineering partner shapes how fast your product moves toward manufacturing, helps shape the product’s mechanical architecture, how much rework you absorb along the way, and how well your product holds up when a supplier asks a question nobody anticipated. The right partner should help reduce avoidable rework, clarify engineering responsibility, and make the path from design to manufacturing easier to manage. Most of these consequences aren’t visible during the sales conversation. They show up three months in, when a revision request takes two weeks longer than expected, or when nobody can explain who actually built the CAD model you’re now trying to hand off to a manufacturer. This article walks through the questions worth asking before you sign anything, not to sell you on one type of engagement, but to help you evaluate any mechanical engineering partner with a clearer picture of what you’re actually buying. A useful way to approach the decision is to focus on what you can verify before signing—not simply what a sales conversation promises. Need a Mechanical Engineering Partner? If you’re assessing partners for a hardware product, start by matching the engineering support to your product stage, team capacity, and manufacturing needs. Explore Engon’s Mechanical Product Engineering Support Short answer: A hardware startup should evaluate an engineering partner based on relevant experience, the actual delivery team, communication structure, commercial transparency, IP ownership, engineering capability, evidence of prior work, and the partner’s ability to support the product as requirements change. The right engineering partner can reduce rework, clarify technical responsibility, and create a smoother path from product concept to manufacturing. Quick Checklist: Before You Sign Before committing to an engineering partner, make sure you can answer these 10 questions clearly: 1. Do they have relevant experience with products like yours? 2. Who will actually work on your project? 3. What exactly is included in the engagement? 4. How will your teams collaborate throughout the project? 5. How will communication, updates, and technical decisions be handled? 6. What will the project realistically cost, including potential additional work? 7. Who owns the CAD files, engineering data, and other project deliverables? 8. What evidence can they provide to support their capabilities and past results? 9. Can they support future revisions, improvements, and sustaining engineering needs? 10. What red flags should you identify before signing? If the answers are clear, specific, and supported by evidence, you’re in a much stronger position to choose a partner that can support your product beyond the initial engineering phase. 1. What Type of Engineering Support Does Your Startup Actually Need? The first mistake many founders make is choosing a vendor before defining the engineering problem. “We just need someone to finish the CAD” often turns out to mean something much broader: mechanical architecture, design refinement, DFM, prototype support, engineering drawings, BOM development, supplier coordination, or manufacturing support, depending on where the product actually stands. Before contacting anyone, it helps to separate a few different engagement shapes: One-time engineering project: a defined, bounded piece of work Engineering team augmentation: added capacity for an existing team Project-based support: ownership of a specific development phase Dedicated engineering team: sustained capacity over a longer arc Ongoing product development: support across multiple product generations Hybrid engagement A mix of the above, depending on how your engineering needs evolve. The clearer these answers are, the easier it is to evaluate proposals based on technical fit, scope, accountability, and overall business value—not simply the lowest price. Questions worth asking internally first: What problem are we actually trying to solve? Do we need additional capacity, or specialized expertise we don’t have at all? Is this a fixed project or an ongoing need? What stays with our internal team no matter who we bring in? Illustrative example: A startup already has an industrial designer and electronics engineer but no mechanical engineering capacity. They probably don’t need a company to take over the entire product; they need a mechanical partner who can pick up the handoff from industrial design, carry it through mechanical architecture, CAD, and DFM, and support the prototype stage. That’s a narrower, more useful engagement than “full product development,” and it’s worth defining before a single call happens. 2. Has the Engineering Partner Worked with Hardware Startups Like Yours? “10+ years of experience” isn’t evidence, it’s a number. What matters is whether the experience is relevant to your product, your manufacturing process, and your stage of development. Worth probing for: Relevant product experience, not just general mechanical engineering Prior work with startups specifically, not only established manufacturers Similar product complexity and similar manufacturing processes Experience carrying a product from prototype to production, not just concept sketches Comfort working with small internal teams and changing requirements Look for evidence that connects the partner’s experience to your actual product stage, manufacturing route, and engineering deliverables. Ask directly: Have you worked with startups before? Can you show relevant examples? Have you taken products beyond CAD and into manufacturing? Have you coordinated with manufacturers directly? Can you explain, specifically, what your team contributed on a past project, not just that you were “involved”? It’s also worth asking about the kind of complexity a partner is used to. A firm that has spent years on large, well-resourced enterprise products may not be the right fit for a startup that needs to make fast decisions with incomplete information, and vice versa, a team used to quick, loosely scoped startup work may struggle once a product needs formal DFM review ahead of tooling. Neither is inherently better; the fit depends on where your product actually is. Short answer: Experience should be demonstrated through specific, relevant evidence, a product, a process, a contribution you can describe, not a years-of-experience figure. Real-world example: From IoT concept to prototype Engon’s IoT enclosure development work provides a relevant example of the type of support a hardware

A hardware product rarely moves directly from a concept sketch to a production line. Between those two points are a series of engineering decisions that determine whether the product will perform as intended, whether its components can be assembled efficiently, whether the selected manufacturing process is appropriate, and whether the design can be produced consistently at the required volume. A CAD model is an important part of that process, but it is only one part. A product can have a complete 3D model and still have problems with material selection, tolerances, assembly access, manufacturability, structural performance, or production economics. Those problems become progressively more expensive to address as development moves from concept to prototype, tooling and production. In practice, many manufacturing problems are not caused by a single major design error. They often come from small decisions made early in development—an inaccessible fastener, an unnecessarily tight tolerance, an unsuitable datum, a difficult-to-machine feature, or geometry that complicates tooling.These seemingly small decisions are often identified during a design review, before they become expensive production changes. This is where hardware product engineering becomes important. Rather than treating CAD, prototyping, DFM, validation and manufacturing support as disconnected activities, product engineering brings them together around the requirements of the finished product. The objective is not simply to create geometry that works in CAD. It is to develop a product whose design, materials, tolerances, assembly method and manufacturing process work together under real production conditions. Hardware product engineering design services help companies move from product concept or existing CAD through mechanical design, prototyping, validation, manufacturability, manufacturing documentation and production readiness. The exact engineering scope depends on the product, development stage, production volume and technical requirements. Almost 70% of manufacturing costs are determined during the design phase and yet, most product teams only begin thinking about manufacturing after the design is finalized. This is a profound problem in product development. Manufacturing Cost Is Often Shaped Early This is why manufacturing considerations need to be introduced early in the product development process. Decisions about materials, tolerances, part geometry, manufacturing processes, assembly methods and tooling can significantly influence the final cost and production feasibility of a product. Addressing these factors during design can help identify potential manufacturing and cost issues before they become expensive changes later in development. Need Engineering Support for Your Hardware Startup? Whether you are refining a concept, developing your first prototype, or preparing an existing design for manufacturing, the right engineering support can help you make better decisions earlier. Explore Engon’s Mechanical Product Engineering Support → What Do Hardware Product Engineering design Services Include? Hardware product engineering can cover different activities depending on the product stage and engineering requirements. Typical services may include: Engineering Service Typical Scope Product architecture Requirements, packaging, interfaces and component relationships Mechanical design 3D CAD, components, assemblies and engineering drawings Prototype development Design iteration, fit checks and functional evaluation DFM/DFA Manufacturability, assembly and production-process review Engineering validation FEA, structural, thermal and functional evaluation where required Enclosure engineering Housing design, PCB integration, sealing, mounting and thermal considerations Manufacturing documentation BOMs, drawings, tolerances, materials and production requirements Production readiness Engineering changes, supplier documentation and manufacturing support This connected approach allows engineering teams to evaluate how an early design decision may affect manufacturing, assembly, validation, cost and future product changes. Who Needs Hardware Product Engineering design Services? Hardware product engineering support can be useful at different stages of product development, including: Hardware startups developing a new physical product Companies turning an early concept into a working prototype Engineering teams that need additional mechanical design capacity Companies moving an existing prototype toward production Manufacturers redesigning a product for a different production process Companies requiring external support for CAD, validation, DFM or manufacturing documentation The required engineering scope should depend on the product’s technical requirements, development stage, production volume and manufacturing objectives rather than on a fixed list of services. 1. Why Product Engineering Decisions Matter Before Manufacturing Product Engineering at a Glance The objective is not simply to create geometry that works in CAD. It is to develop a product whose design, materials, tolerances, assembly method and manufacturing process work together under real production conditions. Requirements Architecture Mechanical Design Prototype Validation DFM Production Product engineering connects early design decisions with the realities of manufacturing, assembly, validation and production. Many of the decisions that influence downstream manufacturing problems are made before a product reaches production. Understanding those relationships early can help prevent expensive changes later. 2. Define the Product Architecture Before Detailed CAD What Needs to Be Defined Early · Product dimensions and packaging constraints · Electronics and component interfaces · Material requirements · Environmental conditions · Expected production volume · Market-specific compliance requirements · Manufacturing processes · Assembly requirements · Serviceability · Performance and testing requirements  These decisions are best considered together rather than in isolation. Early collaboration between industrial design and mechanical engineering can help resolve packaging, usability, structural and manufacturing constraints before detailed CAD begins. Engineering Scenario: Compact Industrial Monitoring Device Consider a compact industrial monitoring device containing a PCB, battery, connectors, and mounting hardware. Before detailed CAD begins, engineers must determine how these elements will fit together, how the electronics will be installed, how the housing will be assembled and serviced, and which manufacturing process will support the expected production volume. These considerations become particularly important for connected hardware and IoT products, where enclosure dimensions, PCB placement, connector access, environmental protection, thermal behaviour and manufacturing requirements can influence one another. For example, an enclosure may need to accommodate the PCB and battery while also providing access to connectors, mounting points and service components. These requirements should be considered before the mechanical architecture is finalized. Validation: Engineers evaluate these requirements together before finalizing the mechanical architecture and detailed CAD. Engineering Insight: Early architectural decisions can affect manufacturing, assembly, serviceability, and downstream component design. Key Takeaway: Defining these relationships early helps prevent costly redesigns later. 3.Design the Product Around Its Manufacturing Process Once the architecture is established, mechanical engineering turns the requirements into physical components and assemblies. This includes 3D CAD, component design,

The cost of mechanical engineering outsourcing for a US company depends on project scope, engineering complexity, and the number of hours required. There is no single fixed price that applies to every project. A simple CAD drafting task will generally require less engineering time than a complete product engineering program involving design for manufacturing, structural analysis, prototyping, and validation. This guide breaks down the mechanical engineering outsourcing cost in the US by pricing model, project scope, and the specific factors that push a budget up or down. It also gives you a practical framework for estimating your own project before you compare vendor quotes. Planning a Mechanical Engineering Project? Before comparing outsourcing rates, it helps to understand exactly what your product requires—from CAD and mechanical design to DFM, prototyping, and engineering validation. Engon Technologies helps hardware startups build the right engineering scope and choose an approach that fits their product stage and budget. Explore Mechanical Product Engineering Services for Hardware Startups → Before comparing outsourcing models or vendor quotes, it is also useful to understand questions to ask a mechanical engineering partner about scope, revisions, ownership and engineering responsibility. What Does Mechanical Engineering Outsourcing Cost in the US? There is no single answer to what mechanical engineering outsourcing costs because the number depends on what is included in the scope. Most firms use one of three common models: hourly, project-based, or dedicated FTE-based support. Hourly pricing works well for smaller tasks, quick design changes, or ongoing support where the scope is not fully defined yet. Project-based pricing works better for defined deliverables, such as a complete product design or a manufacturing-ready CAD package, because it gives the client a fixed number to plan around. FTE-based pricing is suitable for companies that need dedicated engineering capacity over a longer period. Under this model, a company can engage a mechanical engineer or engineering team on a full-time-equivalent basis to work continuously across multiple projects or ongoing engineering tasks. The cost depends on the engineer’s expertise, engagement duration, workload, and required technical skills. Compared with hiring in-house, FTE outsourcing can reduce recruitment, benefits, infrastructure, and other employment-related overheads. The cost of mechanical engineering outsourcing also varies with the type of firm you choose. A solo freelance designer, a boutique engineering studio, and a full-service outsourced engineering team all price differently because they carry different overhead, offer different levels of expertise, and take on different amounts of project risk. In short, there is no single outsourcing price. The right way to think about cost is not “what is the rate” but “what does this specific project actually require.” A well-scoped project with a clear brief almost always costs less to execute than a vague one, because the engineering team spends less time asking questions and redoing work that was never clearly defined in the first place. The sections below walk through exactly what drives that number. Understanding Mechanical Engineering Outsourcing Costs by Engagement Type Mechanical engineering outsourcing is difficult to price with one universal rate because projects vary significantly in complexity and scope. A useful way to compare costs is to separate work into broad engagement types rather than treating every engineering task as equivalent. Because provider rates and project requirements vary, published industry averages should be treated as starting points rather than quotes. For a realistic budget, define the scope first and then estimate the engineering hours, specialist work, prototyping, and revision allowance required for that scope. Several variables combine to shape any mechanical engineering outsourcing pricing model. Understanding each one helps you scope a project accurately before requesting quotes.   Engineering complexity: simple parts cost less than parts with complex surface, assemblies or systems with moving components   Number of engineering hours: this is the single biggest driver of total project cost   CAD and design requirements: 2D drafting is cheaper than full 3D modeling and detailed drawings   Product development stage: early concept work carries more uncertainty than a defined final design   Number of components: more parts mean more design time, more drawings, and more coordination   Revision cycles: each round of feedback and rework adds hours to the project   DFM requirements: design for manufacturing review adds engineering time but reduces production risk   FEA and CAE requirements: structural or thermal analysis requires specialized engineering skills   Prototyping requirements: physical prototypes need engineering support beyond the digital design   Documentation and manufacturing support: complete manufacturing packages take longer to prepare than a basic drawing Mechanical CAD Design vs Complete Product Engineering: What’s the Cost Difference? One of the most common points of confusion for US buyers is the difference between paying for CAD drafting and paying for full product engineering. These are not the same service, and the cost gap between them is significant. CAD outsourcing typically covers 2D and  drafting, basic 3D modeling, individual parts, and engineering drawings based on defined requirements. In this model, the provider is primarily responsible for producing the specified CAD deliverables. Complete product engineering goes further. It may include product development from concept through validation, design for manufacturing and assembly, engineering analysis, prototype support, design iterations, and manufacturing engineering that prepares the design for production. This broader approach is explained in more detail in our guide to hardware product engineering and how engineering decisions connect concept development, validation and manufacturing readiness. If your project only needs drawings for an existing design, CAD-level pricing applies. If your project needs a product built from the ground up and validated for manufacturing and designed with all stakeholders, the scope moves into full product engineering, and the budget should reflect that broader responsibility. Need More Than CAD Support? If your hardware startup needs more than drawings or 3D models, the right engineering partner can support the product through design, DFM, prototyping, validation, and manufacturing readiness. Engon Technologies provides mechanical product engineering support for hardware startups in the US and Europe, helping teams move from product requirements to production-ready designs. Explore Mechanical Product Engineering Support for Hardware

A Design Review is one of the most important checkpoints in any Product Development Process. When engineering, manufacturing, and quality teams sit down to evaluate a design before it moves forward, they have a rare opportunity to catch problems while changes are still cheap. Companies across industries often partner with experts in mechanical product engineering in the US and Europe to validate manufacturability, identify design risks, and ensure products are ready for production before significant investments are made. A thorough Product Design Review can mean the difference between a smooth launch and a production line that grinds to a halt. The cost of finding design issues after tooling or production begins is dramatically higher than catching them during review. A flaw discovered on paper might cost a few hours of rework. The same flaw discovered after a mold has been cut, a fixture built, or a production batch run can cost tens of thousands of dollars and weeks of schedule slip. Effective reviews reduce redesign, delays, and manufacturing risk across the board. This article walks through 10 of the most common design review mistakes companies make and explains how to avoid them, so your next Engineering Design Review can catch critical issues before they reach the shop floor. Overview A Design Review brings cross-functional teams together to evaluate a product before it advances to tooling or production. Common mistakes, such as skipping early-stage reviews, ignoring tolerancing, or overlooking material availability, can delay production, raise engineering costs, and reduce product quality. Structured, cross-functional reviews that include manufacturing, quality, procurement, and supply chain input can improve manufacturability and significantly reduce overall project risk. 1. Skipping or Rushing Early-Stage Design Reviews (PDR/CDR) The Preliminary Design Review (PDR) and Critical Design Review (CDR) are two of the most valuable checkpoints in the Product Development cycle, yet they are often rushed when schedules tighten. The Preliminary Design Review (PDR) confirms that the overall design approach is sound before detailed engineering work begins, while the Critical Design Review (CDR) verifies that the design is ready for release to manufacturing. Skipping or compressing these milestones makes it much harder to identify technical and manufacturing risks before design freeze. Once a design is frozen, the cost of late design changes rises sharply because tooling, fixtures, supplier commitments, and production planning may already be in motion. Early stakeholder alignment during PDR and CDR helps prevent unexpected problems from surfacing downstream, when they are far more expensive and time-consuming to fix. 2. No Cross-Functional Team Involvement A Mechanical Design Review that only includes engineers misses critical perspectives. The strongest reviews bring together Engineering, Manufacturing, Procurement, Quality Assurance, Supply Chain, and Service & Maintenance teams, each contributing a different lens on the same design. A cross-functional design review brings together stakeholders with different responsibilities and decision-making priorities, helping teams identify technical, manufacturing, sourcing, quality, and service risks before they become production problems. Manufacturing flags process limitations before they become production blockers. Procurement identifies sourcing constraints early in the cycle. Quality Assurance raises inspection and tolerance concerns early in the process. Supply Chain flags lead-time and component risk. Service & Maintenance teams highlight long-term serviceability needs. These assembly and serviceability considerations can significantly affect production efficiency and cost; learn more about design decisions that reduce assembly time and manufacturing costs → The benefit of collaborative decision-making is that issues surface while they are still inexpensive to fix, rather than after commitments have already been made. When these perspectives are included early, teams can make better-informed design decisions and reduce the likelihood of costly redesigns, production interruptions, and late-stage changes. 3. Poor Version Control and Engineering Change Management (ECM) When multiple versions of a drawing circulate without clear labeling, different teams end up working from different versions of the design. This creates a serious risk of manufacturing errors, incorrect purchasing decisions, and costly rework. Production might be machining to Rev C while procurement ordered materials against Rev B specs. This mismatch often isn’t caught until parts fail to fit or an inspector flags a discrepancy, by which point time and material are already lost. Engineering Change Orders (ECO) An ECO is the formal instruction to implement an approved change, whether it’s a dimension update, material substitution, or process revision. It typically documents what’s changing, why, who approved it, and the effective date. A well-controlled ECO process ensures that engineering changes are formally reviewed, approved, documented, and communicated before they affect production. Without a proper ECO process, changes get made informally through emails or verbal instructions, which are easy to miss and impossible to audit later. Engineering Change Notices (ECN) While an ECO authorizes a change, an ECN communicates that a change has happened. It’s the notification that pushes updated information out to everyone affected, manufacturing, quality, purchasing, and suppliers, so no one is left working from stale data. This communication step is critical because an approved engineering change is only effective when every affected team receives and implements the updated information. Skipping this step is a common reason outdated drawings keep resurfacing even after a design has already been revised. Drawing management This covers how drawings are stored, labeled, and retrieved so the current revision is always easy to identify. Simple conventions, like consistent revision blocks, clear naming, and restricted edit access, prevent the everyday mix-ups that cause the biggest downstream headaches. A controlled drawing management process should also make it clear which revision is approved for manufacturing and prevent obsolete drawings from being used accidentally. PLM/PDM systems Product Lifecycle Management and Product Data Management systems give this process real teeth by centralizing files, enforcing revision control, and tracking approval history automatically. Rather than relying on individuals to remember which version is current, the system itself controls access and visibility, cutting down on human error significantly. These systems also provide traceability, making it easier to identify when a change was made, who approved it, and which teams or suppliers need to act on the updated information. Preventing outdated files from reaching production The final safeguard is making sure obsolete drawings simply cannot reach the shop