Single-source enrichment fills your data from one provider’s database in a single call. Waterfall enrichment routes each record through an ordered list of providers, then stops the moment one returns a verified value.
So single source is simpler and cheaper per call. Waterfall trades more cost and latency for a higher match rate and broader coverage.
In short, single source is one database, while waterfall is a cascade of many. Most teams that scale outbound end up needing the cascade.
| Dimension | Single-Source Enrichment | Waterfall Enrichment |
|---|---|---|
| What it is | One provider, one database | A cascade of ordered providers |
| Coverage / fill rate | Lower, because of one vendor’s blind spots | Higher, because vendors cover each other’s gaps |
| Match rate | Lower on a single source | Higher across 3 to 4 sources |
| Cost | Lower per call, and predictable | Higher per record, but you pay only for hits |
| Speed / setup | Fast and simple | Slower, with more to configure |
| Best for | Smaller lists, tight budgets, one strong-fit vendor | Scaled outbound, hard-to-reach segments, max coverage |
Both models append the same kinds of fields. Yet they reach a very different match rate.
Let’s start with the short answer. Then we’ll define each term before we compare them head to head.
Waterfall enrichment vs single source: the short answer
Waterfall enrichment beats single source on coverage and match rate, while single source wins on cost and simplicity. A single provider caps your fill rate at its own blind spots.
A waterfall stacks several providers, so each one catches what the last one missed. However, that lift is not free. You pay more per record, you add latency, and you take on more compliance surface.
| If you care most about | Pick |
|---|---|
| Lowest cost and simple setup | Single-source enrichment |
| Highest coverage and match rate | Waterfall enrichment |
So which one fits your team? That depends on your list size, your budget, and how broad your ideal customer profile (ICP) really is. First, though, let’s define waterfall enrichment in plain language, because many readers meet the term for the first time here.
What is waterfall enrichment?
Waterfall enrichment is a method that queries several data providers in order and accepts the first verified value for each field. Picture a real waterfall. A record falls from the first provider to the next, then the next, until a verified value gets caught.
So the name fits the mechanic well. It enriches the contact data your outbound actually needs. That includes work emails, personal emails, and mobile numbers.
It also fills firmographics, which are company traits like size, industry, and revenue. On top of that, it can add technographics, which describe the software a company runs. So one record can leave the cascade far more complete than it entered.
The key idea is simple. No single data provider knows everything. One vendor is strong on US mobile numbers, while another is strong on European work emails.
A waterfall lets each provider do what it does best. As a result, your coverage climbs past what any one source could reach alone. Data vendors describe the same mechanic in their own glossaries, including ZoomInfo’s operations guide to waterfall enrichment.
🧠 Fun Fact: The "waterfall" name borrows from the way water cascades down ledges. Each ledge is a provider, and the record only stops falling once a verified value catches it.
When I built a 3-provider waterfall for a SaaS client in Hamburg’s Hafencity, the logic clicked fast. Our first vendor covered most German firmographics, yet it missed half the mobile numbers.
So we cascaded those misses to two specialists. The mobile match rate jumped, and we still paid the cheap vendor first. That ordering choice mattered more than the vendors themselves.
You now know what a waterfall is. But how does a record actually move through one? Let’s walk the mechanics.
How waterfall enrichment works, step by step
Waterfall enrichment works by sending each record through providers in order and stopping on the first verified hit. It runs per field, so your email and your mobile number can resolve at different providers.
Here’s the flow in plain steps. First, send the record to Provider 1. Then check whether it returns a verified value for the field you need.
If yes, the cascade stops there. If the field is blank, route that miss to Provider 2.
Next, repeat the check. The record keeps falling until a provider returns a verified value or the sources run out.
Because it runs per field, one record can pull its email from Provider 1 and its phone from Provider 3. That is the whole point. You stitch a complete record from many partial ones.
📌 Example: Say you need a VP's work email and mobile. Provider 1 returns a verified email but no phone. So the email locks in, and only the phone request cascades to Provider 2.
Conflict resolution: picking one canonical value
When two providers disagree, you need a rule to land one value. Otherwise you store noise. Most teams resolve conflicts three ways.
First, by precedence, which means a trust order you set in advance. Second, by recency, so the freshest record wins. Third, by verification, so the value that passes email validation or a phone ping wins.
The result is one canonical verified value per field. Deduplication matters here too.
Before you store anything, you merge duplicate records so the same person doesn’t enter your CRM twice. Clean keys make this far easier, which is why normalization comes first.
Why provider ordering is the hidden cost lever
Order your providers cheapest-and-highest-match-for-your-ICP first. This is the single biggest cost lever in a waterfall. Because pricing is pay-per-verified-result, every record a cheap vendor fills is a record you never send to a premium one.
A mistake I made early on was putting our most expensive niche vendor first in the cascade. We paid premium credits to find emails a cheaper provider already had.
After I reordered the cascade cheapest-and-highest-match-first, the cost per valid record dropped sharply. Same coverage, far smaller bill.
So the order is not a detail. It’s the budget.
💡 Pro Tip: Don't order your cascade by brand reputation. Order it by which provider fills the most fields for YOUR ICP at the lowest price. Test that order on a sample before you trust it.
Some tools also lean on models to infer missing fields, an approach you’ll see called AI-assisted enrichment. It can patch gaps, yet it still needs verification before you trust it.
That’s how the cascade resolves a record. Next, let’s see what you give up by using just one provider instead.
What is single-source (single-provider) enrichment?
Single-source enrichment fills your records from one provider’s database through a single API call. One database, one integration, one data processing agreement (DPA), which is the contract that governs how a vendor handles personal data.
So it’s the simplest model you can run. Its strengths are real. Setup is fast, because you wire up one API.
Cost is predictable, because you pay one vendor on one pricing model. Maintenance is light, because there’s only one integration to monitor. For many teams, that simplicity is worth a lot.
The coverage blind-spot problem
Yet single source has one honest limitation. Every provider has coverage blind spots. One vendor might be thin on European mobile numbers, while another lags on early-stage startups or non-tech verticals.
So a single source caps your match rate at that one vendor’s reach, no matter how good it is. You can’t out-configure a coverage gap. You can only add another source.
Back in 2021, most teams just bought one big database and moved on. That worked when lists were small and budgets were tight.
What changed by 2026 was the realization that a single source quietly leaks pipeline through its blind spots. You don’t see the leads you never matched. That invisible miss is the real cost.
Here’s how the blind spot shows up in practice. A US-centric provider might verify 80% of your American contacts yet only 40% of your German ones.
So your reps in the DACH region get half-empty lists, and they assume the market is just harder. It isn’t. Your data source simply can’t see it.
That’s the trap. A single source makes its own gaps look like reality, because you only measure what it returned, not what it missed.
🔍 Did You Know? Even top-tier providers deliver around 97% accuracy on verified records, while the average provider sits closer to 50%, per RocketReach's 2026 benchmark analysis. The gap between vendors is wider than most buyers assume.
So single source is simple but capped, while waterfall is broad but complex. Many readers blur a third option into this picture, though. Let’s clear that up before the head-to-head.
The three enrichment architectures (single-source, parallel multi-source, and sequential waterfall)
There are three enrichment architectures, not two. People often conflate “multi-source” with “waterfall,” but they cost very different amounts. Here’s the clean mental model.
Single-source enrichment queries one provider. Parallel multi-source enrichment queries several providers at once, then merges and deduplicates the results. Sequential waterfall enrichment queries providers in order and stops on the first verified hit.
All three can use the same vendors, yet their economics diverge hard.
Why the billing differs
The difference sits in the billing. Parallel multi-source pays every vendor for every record, even when the first one already had the answer.
Sequential waterfall stops the moment it gets a verified value. So waterfall captures most of parallel’s coverage at a fraction of the cost. That’s the whole reason it became the default for cost-conscious teams.
| Architecture | How it queries | Cost behavior |
|---|---|---|
| Single-source | One provider only | Lowest, one bill |
| Parallel multi-source | All providers at once | Highest, pays everyone every time |
| Sequential waterfall | Providers in order, stop on hit | Middle, pays only until a hit |
📌 Example: Run 1,000 records through three providers. Parallel multi-source bills you for 3,000 lookups. A waterfall might bill 1,000 at Provider 1, then 400 at Provider 2, then 150 at Provider 3.
When I compared the two on a live list, parallel won on raw coverage by a hair. Yet it cost nearly three times more.
The waterfall recovered almost all the same records for a fraction of the spend. For most teams, that trade is obvious. If you’re mapping the broader landscape, plenty of data enrichment tools support one or more of these models.
So those are the architectures. Now let’s put single source and waterfall side by side on the dimensions that decide your bill and your bounce rate.
Waterfall enrichment vs single source: the head-to-head comparison
Here’s the full waterfall enrichment vs single source breakdown. Waterfall enrichment wins on coverage, match rate, and deliverability, while single source wins on cost, speed, and compliance simplicity. The table below goes deeper than the snapshot up top.
Treat every range as a vendor benchmark that shifts by ICP. So run your own pilot before you trust any figure.
| Dimension | Single-Source Enrichment | Waterfall Enrichment |
|---|---|---|
| Coverage / fill rate | Roughly 50% to 70% of a list | Roughly 85% to 95% across providers |
| Raw vs verified match rate | Raw 40-70%, verified 35-60% | Raw 80-95%, verified 70-85% |
| Data accuracy & freshness | Tied to one vendor’s refresh cadence | Cross-checks across vendors, but mixes freshness |
| Cost per valid record | Lower, roughly $0.50 to $2.00 per email | Higher, roughly $1.00 to $3.00 per email |
| Latency / speed | Fast, one call | Slower, one call per hop |
| Integration & maintenance | One API, one mapping | Many APIs, mappings, and failure points |
| Compliance / data provenance | One DPA, one provenance trail | Many DPAs, many provenance questions |
| Bounce rate / deliverability | Higher if records aren’t re-verified | Lower when each hop verifies |
Those coverage and match-rate ranges come from vendor benchmark tables published by Unify, Cleanlist, and Phantombuster in 2026. They are list- and ICP-dependent, so read them as direction, not destiny.
Here’s the single biggest difference. A single source caps your match rate at one vendor’s coverage, full stop. A waterfall stacks coverage from several vendors, so it clears that cap.
That’s the whole reason scaled outbound teams adopt the cascade. But the lift comes with a tax. Each added provider raises your cost per record, adds latency, and adds another compliance trail.
Accuracy and freshness, unpacked
There’s a freshness wrinkle too. A waterfall cross-checks values across vendors, which can flag a stale record. Yet it also mixes data of different ages, so one field might be fresh and another months old.
A single source, by contrast, gives you one consistent refresh cadence. So you trade consistency for coverage. Neither is strictly “more accurate,” because accuracy depends on each vendor’s last verification date, not on the architecture itself.
This is why re-verification beats raw record counts. A vendor advertising hundreds of millions of contacts tells you nothing about whether the 300 records in your niche are current. So ask one question of any provider: when was this field last verified?
So the question isn’t “which is better.” The real question is whether the coverage you’d gain is worth the cost, speed, and complexity you’d pay. Next, let’s price that out properly.
Cost and match-rate trade-offs
The metric that matters is cost per valid record, not cost per call. “More emails found” is the wrong target, because found, verified, and reachable are three different numbers.
A waterfall can find more raw emails yet still raise your bounce rate if those extras fail verification. So separate the three. Found means a provider returned something.
Verified means it passed validation, like an email check or a phone ping. Reachable means the person still works there and the inbox still exists. You only want to pay for verified-and-reachable records.
🔍 Did You Know? A waterfall that "finds" more emails can actually hurt you. If the extra finds aren't verified, they bounce. Hard bounce rates above 5% start damaging your sender reputation, per RocketReach's 2026 deliverability benchmarks.
Diminishing returns and the real budget metric
Now the part most articles skip. Diminishing returns are real. Your second provider recovers a big share of the first’s misses, while your third recovers less.
Each later provider recovers less still. So the last 5% to 10% of coverage costs the most per record. You’re paying premium vendors to find the hardest, rarest records.
That reframes budgeting completely. Don’t ask “how much per lookup.” Ask “how much per valid, reachable record.”
When I ran the same 5,000 records through a single source and then a 4-provider waterfall, the verified match rate jumped nicely. Yet the per-record cost and the latency both climbed too. The fourth provider barely earned its place.
💡 Pro Tip: Measure the marginal lift of each provider. If Provider 4 only recovers 3% more verified records at triple the price, cut it. Coverage for its own sake is a vanity metric.
Why valid records matter beyond the cascade
There’s a wider reason this math matters. Bad data is expensive on its own. IBM’s widely-cited 2016 estimate, published in Harvard Business Review by Thomas Redman, put the true cost of bad data at $3.1 trillion a year for the US economy.
Gartner’s data and analytics research has pegged the average hit to a single organization at around $12.9 million a year. So every dollar you spend chasing unverified records compounds into a much larger waste downstream.
The upside of getting it right is just as large. McKinsey’s analytics research has found that data-driven organizations are 23 times more likely to acquire customers than their peers.
So the goal isn’t the cheapest data. It’s the data your reps can actually trust when they pick up the phone.
So you can see the money trade clearly now. But cost isn’t the only price you pay for a cascade. Speed and compliance carry their own bills.
Speed, complexity, and compliance trade-offs
A waterfall costs you more than money. It costs speed, maintenance effort, and compliance surface. These non-cash costs decide whether a cascade fits your use case, especially for real-time enrichment.
Start with latency, which is the delay before a value comes back. A cascade is only as fast as the providers it must walk through. So real-time enrichment at form-submit gets slower with every hop.
If a prospect is waiting on a “thank you” page, a four-provider cascade can feel sluggish. For batch jobs overnight, latency barely matters. For live lead routing, it can break the experience.
Next, maintenance. More providers mean more API keys, more field mappings, and more failure points to monitor. This is real RevOps debt, the operational drag your revenue operations team carries.
One vendor changes a field name, and your mapping breaks. With one source, that’s one risk. With five, it’s five risks running at once.
The compliance surface scales with every provider
Then compliance, which is the trust focus of this whole topic. One source means one DPA and one provenance trail. So you can answer “where did this data come from” with a single answer.
A waterfall means many DPAs, many lawful-basis questions, and more exposure under GDPR Article 14. That rule covers notifying people when you collect their personal data indirectly, which enrichment usually does.
Running personal data through several vendors means several DPAs. In Germany, GDPR Article 14 notification expectations hit harder than most US teams assume. Every extra provider is another provenance question your legal team will ask.
🔍 Did You Know? GDPR Article 14 specifically governs data you didn't collect directly from the person, which is exactly what enrichment is. So each provider in your cascade adds another lawful-basis question, per the official GDPR text.
US teams face the parallel question under CCPA, which governs how you handle California residents’ personal data. So provenance matters on both sides of the Atlantic.
Data-quality leaders have made this point for years. The named source on data provenance here is the official GDPR text itself:
“Information to be provided where personal data have not been obtained from the data subject.” (GDPR, Article 14)
In plain terms, that covers almost all enrichment, since you rarely collect the data from the person directly. So every provider you add is one more provenance trail to stand behind. The governance work is the real cost of breadth, not the credits.
The compliance surface scales with provider count. That’s a cost too, just one that shows up later as legal review instead of credits.
You’ve now seen every cost a waterfall carries. So when is each model actually the right call?
When single source wins vs when waterfall wins
Single source wins when your list is small, your budget is tight, or one vendor already covers your ICP well. Waterfall wins when you run scaled outbound across regions and verticals one vendor can’t reach.
Here’s the decision framework, broken down by what actually drives the choice.
Pick single source when:
- Your list volume is small, so the coverage gap costs little.
- Your budget is tight, and predictable per-call pricing helps you plan.
- One vendor already matches your ICP well, so a second adds little.
- You need fast, simple setup with one integration.
- Compliance demands a single, clean provenance trail.
Pick a waterfall when:
- You run scaled outbound and every missed match leaks pipeline.
- Your ICP spans regions or verticals no single vendor covers.
- You chase hard-to-reach mobile numbers or personal emails.
- Deliverability matters, so verified-and-reachable data is worth the spend.
- Coverage gaps are quietly costing you more than the cascade would.
A quick lens by team size helps here. A small team sending a few hundred emails a month rarely needs a cascade, because the coverage it would gain is a handful of records. A scaled team running tens of thousands of touches a month feels every coverage gap as lost pipeline, so the cascade pays for itself fast.
ICP breadth tilts it too. A narrow, single-region ICP often sits inside one strong vendor’s sweet spot. A broad ICP spanning regions, company sizes, and verticals will outrun any single vendor’s coverage.
When our team anchored a waterfall on a broad single-source database in 2023 and only cascaded the misses to specialist vendors, the bill fell while coverage held.
That hybrid instinct is worth its own section. So let’s cover it next, because you don’t actually have to pick just one.
The hybrid approach (route by account value)
The hybrid approach enriches your whole list with one affordable single source first, then cascades only the high-value or still-missing records through premium vendors. You don’t have to choose between cheap-but-capped and broad-but-pricey.
You can have most of both. Here’s the logic. Route by value.
Run every record through one affordable broad database. That fills most of your fields cheaply. Then take only the records that still have gaps, or only your highest-value accounts, and cascade those to premium specialists.
So you spend premium credits exactly where they earn their keep. The bulk of your list never touches an expensive vendor.
📌 Example: Enrich 10,000 leads with one affordable source and fill 65% cleanly. Then cascade only the 3,500 misses, plus your top 500 enterprise accounts, to two premium vendors. You capture most of the waterfall's lift without paying premium rates on all 10,000.
This is where a broad single-source database earns a clear role. It works as the affordable anchor, the first node that does the bulk-fill cheaply. The premium vendors then handle only the hard remainder.
As a result, your cost per valid record stays low while your coverage climbs. That’s the best of both architectures in one workflow.
When I set this up for a European SaaS team, the math was stark. The broad anchor handled two-thirds of the list at near-zero marginal cost.
The premium cascade then lifted the hard third. Per-record cost stayed well below a full parallel multi-source run. So the hybrid often beats both pure models on cost-adjusted coverage.
So the hybrid gives you a practical middle path. But how do you actually build one without burning budget on a bad pilot? Let’s get operational.
How to start: building or buying a waterfall the right way
Start by piloting on a real sample and measuring verified match rate, not vendor-claimed coverage. The biggest mistake teams make is trusting a vendor’s “90% coverage” headline.
That number is list-dependent. The only figure that matters is what a pilot returns on YOUR records. Here’s the workflow I use.
First, define the exact field you need per use case. Work email and mobile number are different jobs with different best providers. Second, cleanse before you enrich, so providers match on clean inputs.
Dirty or duplicate keys make every provider miss, so normalization pays off immediately. A focused pass to cleanse before you enrich lifts every downstream match rate.
Third, order providers cheapest-and-highest-match-for-your-ICP first. Fourth, run a two-week pilot on a real sample. Measure verified match rate and cost per valid record, never vendor-claimed coverage.
Fifth, set your dedup and precedence rules so conflicts resolve cleanly. Sixth, re-verify on a cadence, because data decays whether you watch it or not. The Dun & Bradstreet benchmark puts B2B contact decay near 2.1% a month, which compounds to roughly a quarter of your database per year.
💡 Pro Tip: Run your pilot on a sample that mirrors your real ICP mix. A pilot stacked with easy US tech contacts will flatter every vendor and mislead your rollout.
Where a broad single source fits
Where does a tool like CUFinder fit here? CUFinder offers contact enrichment that appends verified fields like email, phone, title, and LinkedIn from a broad database.
Honestly, its honest role in this topic is narrow. It’s a broad single-source database, so it works well as the affordable first node, the anchor, in a waterfall. It is not a turnkey multi-vendor cascade.
Coverage and match rates vary by region and vertical, so test it on a sample first, just as you’d test any provider. No vendor deserves blind trust, including this one.
From here, two next steps help. Compare the best waterfall enrichment tools to pick your cascade vendors. Then follow a build guide on how to set up waterfall enrichment to wire it together.
You’ve got the workflow. Now let’s clear the common traps before they cost you.
Common myths and mistakes to avoid
Most waterfall failures come from chasing the wrong metric or skipping the pilot. Here are the traps I see most often, drawn from real rollouts.
- Chasing 100% fill rate. The last 5% to 10% costs the most per record and rarely pays back. Stop where the marginal cost stops making sense.
- Believing more sources is always better. Each added provider recovers less. A fifth vendor often earns nothing.
- Treating “emails found” as success. Verified and reachable is the real metric. Raw finds that bounce hurt your sender reputation.
- Ordering the cascade by brand. Order by ICP fit and price, not by which vendor has the loudest marketing.
- Set-and-forget with no re-verification. Data decays, so a cascade you built last year is already stale.
- Enriching dirty or duplicate keys. Bad inputs make every provider miss, so clean first.
- Ignoring the compliance surface. Each vendor is another DPA and another provenance question.
- Skipping the pilot. Vendor coverage claims are marketing, not your match rate. Always test on a real sample.
When I skipped a pilot once and trusted a vendor’s coverage claim, the live match rate came in far below the headline. The pilot would have caught it in two weeks. Lesson learned, the hard way.
Now let’s answer the questions readers ask most.
FAQ
What does waterfall enrichment mean?
Waterfall enrichment means querying several data providers in a set order and accepting the first verified value for each field. The record “falls” from one provider to the next until a verified value is caught, then the cascade stops.
So you only pay each provider until a hit lands. It lifts your match rate by letting each vendor cover the gaps the others miss. That’s exactly why scaled outbound teams adopt it.
What is waterfall enrichment on Apollo?
Apollo’s waterfall enrichment is that platform’s version of the general cascade architecture. It routes a record across multiple data sources to lift the match rate, then returns the first verified value it finds.
So treat it as one tool’s implementation, not a different concept. The mechanics match every other waterfall: query providers in order, stop on a verified hit, and resolve conflicts by precedence and verification.
You can read Apollo’s own waterfall enrichment overview for its specific setup. To turn the feature off or stop enrichment, that’s a per-tool setting, so check the tool’s own docs.
What is the difference between single sourcing and multiple sourcing?
Single sourcing pulls data from one provider, while multiple sourcing pulls from several. Single sourcing is simpler, cheaper, and easier to govern, because there’s one database and one DPA.
Multiple sourcing, including waterfall enrichment, raises coverage and match rate by combining vendors that cover each other’s blind spots. The trade-off is cost, latency, and compliance surface. So single sourcing fits small, budget-conscious teams, while multiple sourcing fits scaled outbound.
When compared with single sourcing, what advantage does multiple sourcing have?
Multiple sourcing’s main advantage is higher coverage and a higher match rate, plus more resilience. Because several providers cover each other’s gaps, you match records that any single vendor would miss.
You also gain redundancy, so if one provider’s data goes stale, another can still resolve the field. The cost is real, though.
You pay more per record and take on more providers to manage. For hard-to-reach segments, that lift usually justifies the spend.
What is an example of single sourcing?
A clear example of single sourcing is enriching your whole list from one provider’s database through one API. You upload your records, the provider appends emails, phones, and firmographics, and you’re done.
There’s one contract, one integration, and one provenance trail. Many teams start here because it’s fast and predictable. The limit shows up as coverage gaps in regions or verticals that vendor doesn’t cover well.
What are the best waterfall enrichment tools?
The best waterfall enrichment tools depend on your ICP, but several platforms now offer the cascade natively. Options include Apollo, Clay, BetterContact, and others.
The right pick balances coverage for your segment, per-record cost, and how easily it integrates with your CRM. So rather than chase a single “best,” test two or three on your own sample. For a full breakdown, see our guide to the best waterfall enrichment tools linked above.
How do you measure if a waterfall beats your single source?
Run a two-week pilot on the same real sample through both setups, then compare verified match rate and cost per valid record. Don’t compare vendor-claimed coverage, because those numbers are list-dependent marketing.
Send the same 2,000 to 5,000 records through your single source and through the cascade. Then count how many records each one verified, and divide your spend by those verified records. The cascade wins only if its extra coverage justifies its extra cost on YOUR data.
Is waterfall enrichment compliant under GDPR and CCPA?
Waterfall enrichment can be compliant, but each provider adds a DPA and a lawful-basis question. Under GDPR, Article 14 applies because enrichment collects personal data indirectly, so notification expectations come into play.
Under CCPA, you must handle California residents’ data per their disclosure and opt-out rules. So the architecture isn’t inherently non-compliant.
The work is governance: track provenance per provider, sign a DPA with each, and document your lawful basis. More providers simply means more of that work.
The bottom line
Single source and waterfall enrichment aren’t really rivals. They’re two points on a cost-and-coverage curve. Single source is one database: cheap, fast, and capped at one vendor’s blind spots.
Waterfall is a cascade of many: broader and higher-matching, but pricier, slower, and heavier to govern. So start with the simplest source that covers your ICP.
Then add a cascade once coverage gaps start costing you real pipeline. And whatever you do, validate the choice on a two-week pilot, because the only match rate that matters is the one your own records return.
Bad
Okay
Good
Amazing
