Beverage recipe development is the process of turning a drink idea into a working, commercially viable prototype. Most beverage businesses jump straight to manufacturing without completing a structured beverage recipe development process — and it is exactly why so many drink launches fail at shelf. Before you can formulate a drink for production, you need a recipe that actually works. That is what professional beverage recipe development services deliver: a tested, sensory-validated recipe that can survive commercial scale-up without expensive mid-project reformulation. This guide walks through the complete beverage recipe development journey — from concept to approved prototype.
What Is Beverage Recipe Development and Why Is It Important?
Beverage recipe development is the process of creating a working drink formulation from a product concept — defining what goes into the drink, in what proportions, and achieving a sensory outcome (taste, colour, aroma, mouthfeel) that meets both consumer expectations and commercial production requirements.
It is the creative and technical phase that happens before commercial formulation. A recipe is not yet a formula. A recipe defines what you want the drink to be. A formula defines how to make it at scale, under process conditions, with shelf-stable ingredients, at target cost.
Recipe development answers the questions: What does this drink taste like? What is in it? Does it work sensory? Is it buildable?
Beverage recipe formulation — which is a different and later stage — answers: Can we make this consistently at commercial volume? What are the process parameters? What is the declared shelf life?
The distinction matters because rushing from concept to formula without proper recipe development produces products that fail sensory testing, require expensive reformulation mid-project, or reach retail with a taste profile that misses the target market entirely.
In practice: Beverage recipe development typically requires 4–12 prototype iterations before a recipe is ready to proceed to the beverage formulation process. Brands that skip this stage almost always revisit it — expensively — after production trials.
Beverage Recipe Development vs Beverage Formulation: What’s the Difference?
These two terms are frequently used interchangeably in the industry, and that causes expensive confusion. Here is a clear distinction.
| Aspect | Beverage Recipe Development | Beverage Formulation |
|---|---|---|
| Stage | Concept → working prototype | Prototype → manufacturable specification |
| Goal | Create a drink that tastes right | Create a formula that can be made at scale |
| Environment | Kitchen / bench-scale lab | Pilot plant / manufacturing scale |
| Key outputs | Working recipe, sensory benchmarks | Manufacturing spec, process parameters, shelf life data |
| Ingredients used | May include fresh/perishable ingredients | Commercial shelf-stable equivalents |
| Shelf life focus | Not primary concern | Central technical requirement |
| Cost focus | Getting the taste right | Bill of materials at target COG |
| Regulatory output | None at this stage | Nutrition facts, ingredient declarations |
| Who leads it | Food technologist + sensory team | Food technologist + process engineer |
| What comes next | Formulation begins | Pilot trials, stability testing |
Understanding this distinction is important when working with a beverage formulation partner. Recipe development and formulation are usually billed and scoped separately — because they require different skills, environments, and timelines.
Step-by-Step Beverage Recipe Development Process
There is no single formula for developing a beverage recipe — but there is a logical sequence of decisions that separates structured development from expensive guesswork. Here is the process FFCE’s beverage R&D team follows across all beverage product development projects.
1
Define the Product Concept
Before a single ingredient is selected, the concept must be clearly defined: What type of drink is this? What consumer occasion does it serve? What is the brand positioning — premium, functional, natural, nostalgic, indulgent? A poorly defined concept produces a directionless recipe. Spend time here. Every decision downstream flows from this definition.
2
Define the Target Consumer
The consumer drives sensory expectations. An energy drink targeting 18–25 year old male fitness consumers requires a very different flavour profile, sweetness level, and mouthfeel than a wellness herbal beverage targeting 35–50 year old female health consumers. Sweetness tolerance, acidity preference, and flavour familiarity all vary significantly across demographics. Define your consumer before you define your recipe.
3
Select the Beverage Category
The category determines the technical framework: pH targets, Brix range, carbonation level, packaging format, and preservation system. A carbonated soft drink, a functional RTD, and a plant-based milk all require fundamentally different recipe architectures. Selecting the category is not just a marketing decision — it is a technical starting point.
4
Select Base Ingredients
Water quality is the foundation. Hard water affects taste, soft water affects mouthfeel. Most beverage recipes require treated or RO-purified water with controlled mineral content. After water: your primary base — juice, tea extract, dairy, plant milk, or plain water for carbonates. The base determines everything that follows.
5
Sweetener Selection and Brix Targeting
Sweetness is the most technically nuanced decision in structured beverage R&D. Sucrose gives clean sweetness with a fast onset. HFCS gives sweetness with a slight delay. Steviol glycosides give intense sweetness with a lingering aftertaste that must be masked. Erythritol gives cooling mouthfeel. Each sweetener behaves differently across pH, temperature, and processing conditions. Target Brix based on category benchmarks, then select sweetener system to achieve it.
6
Acid Balance and pH
Acidulants do three things in a beverage recipe: they adjust pH for microbial safety, they contribute to flavour perception (tartness), and they modulate sweetness. Citric acid is the most common. Malic acid gives a softer, more fruit-like tartness that lingers pleasantly. Phosphoric acid is used in colas. The ratio of sweetness to acidity — sometimes expressed as Brix:acid ratio — is what determines whether a drink tastes balanced or cloyingly sweet or harshly sour. Getting this ratio right takes several prototype iterations.
7
Flavour System Development
Flavour selection for a beverage is not picking a flavour compound from a supplier catalogue. It is selecting a primary flavour, a top note, a middle note, and sometimes a masking agent — then testing how each interacts with your base, sweetener, and acid system at the target pH and temperature. Flavour performance changes significantly between ambient and chilled, between carbonated and still. Benchmark against 2–3 competitor products. Know what you are matching and where you are differentiating.
8
Mouthfeel Engineering
Mouthfeel is the most underestimated element of developing a beverage recipe. Consumers describe mouthfeel as “thin,” “watery,” “thick,” “creamy,” “smooth,” or “heavy” — and it is driven by viscosity, dissolved solids, carbonation level, and fat content. In zero-sugar beverages, reduced Brix creates thin mouthfeel that feels watered-down. Hydrocolloids (guar gum, pectin, carrageenan at very low use levels) can restore body. Carbonation level is also a mouthfeel decision as much as it is a category decision.
9
Colour Development
Colour sets consumer expectation before the first sip. A watermelon drink that is pale yellow will fail sensory before it is tasted. Colour can come from natural sources (anthocyanins, beta-carotene, turmeric, spirulina) or from approved synthetic colours depending on market and positioning. Natural colours are significantly more technically challenging — they are pH-sensitive, light-sensitive, and heat-sensitive. If your product brief says “natural colours,” build colour stability testing into your recipe development timeline from the start.
10
First Prototype Development
The first bench prototype is a starting point, not a finished recipe. Mix your selected ingredients at bench scale (typically 1–5 litres). Evaluate immediately for colour, aroma, initial sweetness, acid balance, and mouthfeel. Do not expect to nail it on the first iteration — experienced beverage R&D teams typically plan for 6–12 iterations. The first prototype establishes the starting point for systematic refinement.
11
Sensory Evaluation
Sensory evaluation of a beverage prototype should be structured, not casual. Use a consistent panel (minimum 5–8 trained tasters for internal evaluation), evaluate at the correct serving temperature (chilled for RTDs, ambient for shelf-stable), and use a standardised evaluation scorecard covering: appearance, aroma, initial taste, mid-palate, finish, and overall balance. Document every session. The documentation becomes your reformulation guide.
12
Recipe Refinement and Iteration
Recipe refinement is systematic, not random. Change one variable at a time. Increase sweetness by 0.5 Brix and re-evaluate. Increase acid by 0.05% and re-evaluate. Change flavour usage level by 10% and re-evaluate. Changing multiple variables simultaneously makes it impossible to identify which change produced the improvement — or introduced a new problem. Maintain a structured iteration log. This discipline separates structured, professional recipe work from intuition-based guessing.
Expert Tip — FFCE Beverage R&D Team
Most successful commercial beverages go through 6–12 recipe iterations before they are ready for formulation. Document every ingredient change and sensory observation during development — this record becomes invaluable during commercial scale-up, investor due diligence, and any reformulation work that comes later. Brands that skip this documentation consistently spend more time reformulating than brands that record it from day one.
Beverage Categories and Their Recipe Development Characteristics
Every beverage category presents specific technical challenges in recipe development. Understanding these characteristics helps you plan development timelines, ingredient strategies, and testing requirements accurately.
🥤
Carbonated Drinks
CSDs, sparkling waters, sodas
🍊
Juice Beverages
NFC, from concentrate, nectars
⚡
Energy Drinks
Caffeine, taurine, B-vitamins
🏃
Sports Drinks
Electrolytes, carbohydrates
🌿
Functional Beverages
Adaptogens, collagen, probiotics
🥛
Dairy Beverages
Flavoured milk, lassi, kefir
🌱
Plant-Based
Oat milk, almond milk, rice drink
🍵
Tea Beverages
RTD tea, kombucha, cold brew
☕
Coffee Beverages
RTD coffee, cold brew, coffee milk
🌸
Herbal Drinks
Hibiscus, tulsi, ashwagandha shots
Category-Specific Considerations
Carbonated drinks require gas retention as a key recipe parameter — carbonation level (volumes of CO₂) directly affects perceived sweetness and mouthfeel. Higher carbonation suppresses sweetness, requiring Brix adjustment.
Juice beverages have natural flavour variability between batches and seasons. Recipe development must establish sensory target ranges — not exact values — to accommodate this natural variation while maintaining a consistent consumer experience.
Energy drinks have complex flavour masking requirements. Caffeine, taurine, and B-vitamins all contribute bitterness or off-notes that must be systematically masked in the recipe without over-sweetening.
Functional beverages present some of the most technically demanding recipe development challenges. Botanical extracts, adaptogens, and collagen hydrolysates all contribute taste and aroma characters that may conflict with the primary flavour concept. Dosage, solubility, and taste masking must be resolved at recipe stage before proceeding to the formulation process.
Plant-based beverages have specific mouthfeel challenges — plant proteins create chalky, astringent mouthfeel at higher concentrations. Emulsification, homogenisation parameters, and stabiliser selection are critical recipe decisions that must be benchmarked at prototype stage.
Ingredient Selection for Beverage Recipe Development
Ingredient selection is where recipe development becomes highly technical. Each ingredient category carries specific functional, sensory, regulatory, and cost implications. Here is how to think through each category.
Water
The foundation of every beverage. Mineral content, chlorine presence, and hardness directly affect taste, colour stability, and ingredient solubility. Most commercial beverage recipes require reverse osmosis (RO) water with controlled remineralisation.
RO water, RO + mineral blend, filtered municipal water
Sweeteners
The most complex ingredient decision. Sweetener selection affects Brix, aftertaste profile, mouthfeel, calorie content, regulatory labelling, and consumer positioning. Blending multiple sweeteners is common in zero-sugar formulation to balance onset and finish.
Sucrose, glucose, HFCS, stevia, erythritol, monk fruit, sucralose, aspartame
Acids (Acidulants)
Acidulants control pH for microbial safety, contribute tartness and brightness to flavour, and modulate the perception of sweetness. The Brix:acid ratio is the single most important balance point in carbonated and juice beverage recipes.
Citric acid, malic acid, tartaric acid, phosphoric acid, lactic acid, ascorbic acid
Flavours
Flavour systems for beverages are complex blends of aroma compounds, not single molecules. Natural flavours, artificial flavours, and WONF (with other natural flavours) all carry specific regulatory implications. Usage level is critical — too much is as problematic as too little.
Natural fruit flavours, botanical extracts, essential oils, WONF flavour systems
Colours
Colour stability across pH, heat, and light exposure varies significantly between natural and synthetic colour systems. Natural colours (anthocyanins, beta-carotene, spirulina) are pH-sensitive. Colouring foodstuffs (Fruit Juice, Vegetable Juice) carry different regulatory requirements than food colours.
Anthocyanin, beta-carotene, turmeric, spirulina, caramel colour, FD&C certified colours
Preservatives
Preservation approach is determined by pH, packaging, and processing. High-acid beverages (pH below 4.0) have inherent microbial safety at ambient. pH above 4.0 requires either thermal processing, chemical preservation, or cold chain. “Natural” positioning limits chemical preservative options significantly.
Sodium benzoate, potassium sorbate, natamycin, rosemary extract (natural)
Stabilisers and Hydrocolloids
Stabilisers prevent phase separation, control viscosity, and deliver mouthfeel. In juice beverages, stabilisers suspend pulp. In plant-based milks, they prevent cream-line and sedimentation. Usage levels are typically very low (0.01–0.5%) but have significant sensory impact.
Carrageenan, guar gum, xanthan gum, pectin, cellulose gum, acacia gum
Vitamins and Minerals
Vitamin and mineral addition drives functional claims but introduces taste challenges. B-vitamins (particularly B3, B6, B12) contribute bitterness and off-notes at therapeutic dose levels. Mineral salts (sodium, potassium, magnesium) add electrolyte positioning but affect flavour. Recipe development must address masking requirements.
Vitamin C, B-complex, Vitamin D, magnesium citrate, potassium chloride, zinc
Botanical Extracts
Botanical ingredients (adaptogens, herbs, plant extracts) are the fastest-growing addition to functional beverage recipes. Most botanicals have strong, complex flavour characters that compete with the primary flavour concept. Solubility, stability, and regulatory status must all be confirmed at recipe stage.
Ashwagandha, tulsi, ginger, turmeric, hibiscus, green tea extract, elderberry
Caffeine and Active Ingredients
Caffeine, taurine, guarana, and L-theanine all require flavour masking at effective dose levels. The bitterness of caffeine at 80–150mg per 250ml serving must be addressed in recipe development — not left for formulation. Regulatory dose limits vary by market.
Caffeine, taurine, L-theanine, guarana extract, ginseng extract, creatine
Beverage Flavour Development: The 6 Sensory Dimensions
Flavour development for a beverage recipe is not a single decision — it is the management of six interdependent sensory dimensions, each of which affects the others. A technically excellent recipe achieves balance across all six simultaneously.
1. Sweetness
Sweetness is the anchor of most beverage recipes. It is perceived in the first half-second of contact with the palate and sets the baseline expectation for everything that follows. Sweetness level is expressed in Brix (for sugar-sweetened beverages) or modelled by equivalent sweetness intensity (for reduced-sugar systems). The goal in recipe development is finding the sweetness level that feels satisfying but not cloying — typically benchmarked against the 2–3 leading competitors in your category.
2. Acidity
Acidity is what gives a beverage brightness, freshness, and liveliness. A beverage without sufficient acidity tastes flat and syrupy even at moderate Brix. The Brix:acid ratio — typically 14–20:1 for fruit beverages — is the starting point. Acid type matters as much as level: citric acid gives sharp, immediate tartness that fades quickly; malic acid gives a softer, rounder tartness that lingers. For a more complex flavour profile, blending acid types is common.
3. Aroma
Aroma contributes approximately 70–80% of what we perceive as flavour. In a beverage recipe, aroma comes primarily from the flavour system, with secondary contributions from botanical extracts, fruit juices, and tea or coffee bases. Aroma volatility matters: head-space aroma (what you smell on opening) is driven by high-volatility top notes; retro-nasal aroma (what you experience while drinking) is driven by mid and base notes. Both need to be addressed in flavour selection.
4. Mouthfeel
Mouthfeel is the physical sensation of a beverage — viscosity, body, creaminess, carbonation tingle, astringency, or oiliness. It is separate from taste but powerfully affects overall perception of quality. A thin mouthfeel reads as “watered down” even when Brix is correct. Recipe development should include explicit mouthfeel targets alongside taste targets, and address mouthfeel through sweetener system, dissolved solids, carbonation level, and hydrocolloid addition as needed.
5. Aftertaste
Aftertaste is what lingers on the palate 30–60 seconds after swallowing. A clean finish is desirable in most beverage categories. Stevia-based sweetener systems are notorious for a lingering sweet aftertaste or a slightly bitter metallic note that consumers often describe as artificial. High caffeine levels leave bitterness. Some botanical extracts leave herbal or earthy aftertaste. Addressing aftertaste in recipe development often requires masking agents, sweetener system reformulation, or flavour system adjustments.
6. Balance
Balance is the overall harmony of all five dimensions above. A balanced beverage is one where no single dimension dominates — where sweetness, acidity, aroma, mouthfeel, and finish work together to produce a cohesive sensory experience. Balance cannot be engineered in isolation. It emerges from the iterative refinement process described earlier. Experienced beverage recipe developers know what “balanced” feels like intuitively — and can identify which dimension is disrupting balance from a single sip.
Beverage Prototype Development Process
Prototype development is where recipe decisions become physical reality. The first prototype is never the final prototype — and experienced beverage development services teams plan for this explicitly.
At bench scale, a prototype batch is typically 1–5 litres. This allows ingredient adjustments to be made in small increments at low cost. Each prototype iteration should test one or two variables — never rewriting the entire recipe at once. The iteration log — recording what changed, what the sensory result was, and what the next adjustment will be — is the most valuable document produced during recipe development.
Temperature matters significantly in prototype evaluation. Most finished beverages are consumed chilled. Evaluating a prototype at ambient temperature gives a different sweetness perception (sweetness decreases at lower temperatures) and different carbonation sensation. Always evaluate prototypes at their intended serving temperature.
Prototype stability also needs to be assessed even at recipe stage. Natural colours and flavours can shift significantly over 24–48 hours in a prototype — and this early stability assessment often reveals technical issues that would otherwise emerge at the formulation stage at much higher cost.
Beverage Recipe Testing and Validation
Recipe testing is not a single event — it is a structured sequence of evaluations that progressively increases confidence in the recipe before committing it to the formulation stage.
Internal Sensory Testing
Internal tasting panels (5–10 trained tasters) provide fast, low-cost feedback during early iterations. The panel should include tasters with calibrated palates who can articulate specific sensory attributes — not just “I like it” or “it needs more flavour.” Use a structured scorecard evaluating appearance, aroma, sweetness, acidity, mouthfeel, balance, and overall acceptability on a defined scale.
Blind Benchmarking
Benchmark your prototype blind against 2–3 leading competitor products in your category. This is one of the most valuable exercises in beverage R&D. It reveals where your prototype genuinely stands in the competitive landscape — not where you think it stands. Blind benchmarking often reveals that the “close enough” prototype is actually meaningfully inferior to the market leader in one specific dimension.
Consumer Testing
Qualified consumer testing with your actual target demographic should occur when the recipe is close to final — typically after 3–4 internal iterations have narrowed the sensory profile. Consumer testing at this stage validates that the recipe resonates with real consumers, not just the development team. A small group of 20–30 consumers in a structured tasting session provides statistically meaningful direction.
Reformulation
Consumer testing results should be used to define the final refinements. The goal is not to satisfy every consumer’s individual preference — it is to find the recipe that performs well on the core sensory attributes that drive category preference in your target demographic. Some consumer feedback will contradict other feedback. The food technologist’s role is to interpret the pattern and direct the final refinement accordingly.
Common Beverage Recipe Development Mistakes
These are the mistakes FFCE’s beverage R&D team sees most frequently when brands come to us having already invested in recipe development that did not reach a viable prototype.
- Developing the recipe without defining the category clearlyA functional water and a sports drink are not the same category even if both contain electrolytes. The category determines pH, Brix, carbonation level, preservation approach, and regulatory requirements. Starting without a clear category definition produces a recipe that does not fit any market.
- Selecting sweetener system based on cost aloneThe cheapest sweetener system almost never produces the best sensory outcome — and in zero-sugar formulation, sweetener selection is one of the most complex decisions in the project. Brands that anchor to sucralose or stevia purely for cost reasons typically encounter aftertaste problems that require complete sweetener system reformulation later.
- Evaluating prototypes at the wrong temperatureSweetness decreases significantly at lower temperatures. A recipe evaluated at room temperature will taste noticeably different when chilled. Always evaluate beverages at their intended serving temperature — this is a basic but frequently overlooked discipline.
- Changing multiple variables between iterationsChanging sweetener concentration, acid level, and flavour usage level simultaneously in one iteration makes it impossible to identify which change produced the improvement — or introduced a new problem. Change one variable at a time. Document every iteration. This is non-negotiable in structured recipe development.
- Not addressing mouthfeel explicitlyMouthfeel is consistently the most overlooked dimension in recipe development briefs. Brands specify taste targets in detail but leave mouthfeel undefined. The result is a recipe that tastes right but feels thin, watery, or grainy — and these problems are significantly harder to fix at formulation stage than at recipe stage.
- Using fresh or perishable ingredients in the recipe without planning their commercial replacementFresh ginger, real lime juice, and fresh herbs produce excellent recipe prototypes. But their commercial replacements (extracts, concentrates, flavours) never taste identical. Building a recipe around fresh ingredients without confirming commercial ingredient availability is planning to reformulate expensively later.
- Skipping blind benchmarkingThe team that developed the recipe cannot objectively evaluate it against competitors — familiarity bias is too strong. Blind benchmarking against market leaders is the only honest way to assess where the prototype genuinely sits in the competitive landscape.
- Ignoring natural colour stabilityNatural colours — particularly anthocyanins and chlorophyll derivatives — are pH-sensitive and can shift dramatically in colour between a fresh prototype and a product that has been stored for even a few weeks. Discovering colour instability at formulation or production stage is expensive. Test natural colour stability early at recipe development stage.
- Relying on the internal team for all sensory evaluationSensory adaptation (palate fatigue) is real, and so is confirmation bias. The team that has been tasting the prototype daily for three weeks will evaluate it very differently from a consumer tasting it for the first time. Bring in external tasters — even informally — at regular intervals during development.
- Not confirming ingredient regulatory status before finalising the recipeSome botanical extracts, flavour compounds, and functional ingredients are not approved for use in beverages in all markets. Discovering a regulatory non-compliance after the recipe is finalised requires reformulation from scratch. Confirm regulatory status of every functional ingredient at the recipe development stage.
- Proceeding to formulation with a recipe the team internally “likes” but has not consumer-testedInternal approval does not equal consumer acceptance. A recipe that the development team finds “pleasantly balanced” can be “too mild,” “too sour,” or “too unusual” for the target consumer. Consumer validation — even at a small sample size — before formulation commitment is an investment that consistently prevents expensive later-stage reformulation.
DIY vs Professional Beverage Recipe Development
Many beverage brands start with DIY recipe development — kitchen experiments, home-scale prototypes, and informal tasting sessions. Here is an honest comparison of what each approach delivers.
| Aspect | DIY Recipe Development | Professional Beverage R&D |
|---|---|---|
| Cost | Lower upfront cost | Higher upfront, lower total project cost |
| Timeline to viable prototype | Highly variable — often 6–18 months | Structured 6–14 week timeline |
| Ingredient knowledge | Limited to familiar ingredients | Access to full commercial ingredient portfolio |
| Sensory evaluation | Informal, biased, undocumented | Structured, calibrated, documented |
| Commercial viability | Often requires major reformulation for production | Designed for commercial production from the start |
| Regulatory awareness | Usually low until it becomes a problem | Regulatory requirements built into recipe from concept |
| Iteration discipline | Often unfocused — too many variables changed at once | One variable per iteration, fully documented |
| Shelf life awareness | Not typically assessed at recipe stage | Early stability assessment built in |
| Consumer testing | Often informal or skipped | Structured consumer panel at defined milestones |
| Transfer to formulation | Often problematic — recipe needs significant rework | Smooth handoff — recipe designed for formulation from start |
DIY recipe development is not inherently wrong — many successful beverage brands started with kitchen experiments. The critical transition point is when you decide to take the product to commercial production. At that point, professional recipe development support ensures that what you have built can actually be manufactured, stabilised, and delivered to consumers consistently at scale.
What Happens After Beverage Recipe Development?
A completed beverage recipe — a working prototype with documented sensory targets, ingredient list, and approximate quantities — is ready for the next stage: beverage recipe formulation.
At the formulation stage, each recipe ingredient is specified in commercial form (not fresh/bulk/artisanal equivalents), processing conditions are defined (pasteurisation temperature, homogenisation pressure, carbonation volume), and the formula is scaled from bench to pilot to commercial volume. This is covered in detail in our guide to beverage recipe formulation.
After formulation, the product enters stability testing — accelerated and real-time shelf-life studies that validate the declared shelf life at the target storage conditions. Colour, taste, aroma, microbiology, and nutritional values are all monitored over the study period. This is a non-negotiable step before any commercial launch and is covered in detail in our beverage formulation process guide.
Converting your recipe into a manufacturing-ready commercial formulaAfter FormulationBeverage Formulation ProcessStep-by-step: from formula to commercial productionOverviewBeverage Formulation (Pillar)Complete guide to the beverage development journey
Written by FFCE Beverage R&D Team
Food Technologists, Beverage Formulators & Sensory Specialists · 13+ Years Experience
Written by the FFCE beverage team, who have developed recipes across 100+ commercial beverage projects — carbonated drinks, energy drinks, RTD teas, plant-based milks, functional shots, and dairy beverages. We have seen enough failed prototypes to know exactly what goes wrong, and exactly how to prevent it.

